Customer Experience

What about the customers v2 part 2

NPS (36) – Sustained Leadership Commitment – How to get and keep it – 36th article in a series on the Net Promoter System®

/ , , / , , , , /

Welcome to the 36th article in my series on the Net Promoter Score and System. This time the subject is a tricky one: how to get and keep leadership commitment to customer experience work. These techniques have worked for me. I would love to hear what has worked for you.

You can read all of the prior articles in the NPS series on our blog page here.

Most leaders have a detailed understanding of the area they are responsible for. The majority do not have as detailed an understanding of what they need to improve for customers. This is where the Net Promoter System comes in. If you lead Customer Experience, this should be your main objective. At HP, I made it my business to propose customer-centric priorities to each of the other leadership team members as they went through their annual strategy cycles. At least some of these made it to their formal priority lists. I admit going through a ‘name and shame’ cycle with screen captures of Intranet sites that listed each leader’s priorities. I also supplied suggestions to the people who put those intranet sites together.

Short versus medium-term

Customers take time to notice and provide feedback on improved customer experience. Improvement actions you take today may take 6 to 18 months to produce a measurable difference in competitive NPS benchmarks. In the meantime, your leaders may have pressure to produce revenue and cost numbers each quarter that cause them to do things that negatively impact customers. This is the one area where there is no real solution. Almost nothing you do in customer experience this week can impact your company results this fiscal quarter. Many other functions have the same challenges, notably Marketing.

Trying to get leaders to act on facts, not intuition

As my own team published the quarterly competitive NPS benchmarks, I asked to be on the management team agenda for each business and function. The purpose was to review results and priorities. Without exception, the teams were highly engaged. There is somewhat of an art to such presentations, as it is all too easy to show up for a 45-minute presentation slot with 50 slides. Try to prepare the discussion with one slide per 7 to 10 minutes of discussion. You can add backup slides for distribution. The audience should be at least as interested in your insights about competitors’ trends as about your own.

Facing reality — leaders are still going to act first on the basis of intuition and emotion

Feed the intuition. Use emotion. Start your presentations, discussions, speeches and emails using an appropriate emotional appeal to your audience’s intuition. Dan Ariely refers to closeness, vividness and the ‘drop-in-the-bucket’ effect in The Upside of Irrationality.

  • Closeness is both physical and psychological. If you happen to be in New Zealand, start with an example of a problem faced by a specific local customer. Do not start by talking about a customer in India. Pick a problem your audience can identify with from their everyday lives, even if you believe it is not the most important problem. Think about using long queues for something, the prized Christmas toy being perpetually out of stock at the local store, or perhaps an error in a credit card charge. The problem does not have to be exactly the same as the one on the ‘Top 5 customer problem’ list; just close enough for your audience to relate to it personally.
  • Vividness is about including emotion in the way you tell the story. Let’s suppose your customers are complaining about delivery errors. Your story could be about how you ordered the latest popular children’s toy online for your nine-year-old daughter. You were delighted when the box showed up the day before her birthday, gift-wrapped like you asked. Then tell the audience how your daughter cried when she opened the box and the wrong toy was inside, instead of the long-promised gift.
  • Drop-in-the-bucket is about whether the people in the audience feel that it is within their individual power to make a difference. If they feel they could work hard on something that would take a long time and make a marginal difference, they will lose interest. If you are discussing a big and difficult problem, you should represent the problem you are addressing in terms of named individuals who work for local customers. At least the audience will feel a desire to help their local ‘drops in the bucket’. It is far more effective to sub-divide a problem into smaller components that provide benefits in the short term. If you are showing that a competitor has a clear lead in three areas, pick one (or a subset of one) that the audience you are addressing can recognize and address themselves. This of course means changing what you say and write, adjusting it to each new audience.

Customer ‘Open Hour’ at Microsoft North America

When Robert Youngjohns was head of Microsoft North America, he used to do something I admire. He held an ‘Open hour’ for customers about once a month. What this meant was that all customers knew they could dial in to a conference call line and talk to Robert. Since there was no practical limit to the number of people who could join, customers could also hear and dialog with each other. If you copy that, your executive needs to have a reasonably complete grasp of the business and current known issues, and have the self-confidence necessary to say, “I don’t know, and will find out” when necessary.

Executive sponsor programs

A practical way of letting senior people spend more time with customers is to implement an executive sponsor program. The role of the sponsor is to improve the overall relationship between your company and the customer or customers to which they are assigned. You should start by agreeing assignments to a single customer each. Ideally, the assignments should correspond to each executive’s areas of personal interest. Assigning them to former employers also works, provided they separated on good terms.

It is hard for executives to say no to the program, but not all will actually become active. It can be hard at the customer end too, as other companies may be trying to get time with the same contacts. Somehow, your program has to be different or better than your competitors’, or you just won’t get the engagement you require. Note that sponsors should never start off by participating in sales discussions about a current deal. That would destroy their relationship-centric credibility.

Listen to calls

Please note the plural in “calls”. It is a good idea to let senior people listen to multiple calls in your service center. Once they understand the agents’ work, you can let them take a call, supporting them in the same way you would support a new call center agent. If executives listen to calls, it is important to select the calls at random. If they only listen to your best-performing agent, they will have a biased view of the work. If your executives do not listen to enough calls, or just do it once, for a few hours, they will walk away believing that what they have heard is entirely representative of customer views. This means that the customer experience leader needs to do an excellent job of positioning the service center work as a hygiene factor, and needs to remind the executive that most customers do not call the service center. Unfortunately, many large companies outsource service center work for cost reasons, making executive participation difficult to arrange.

Next time

The next article will be the last in the series. I will suggest how you should start the NPS journey and make some closing remarks. And there will be a quiz!

 

As is often the case, the above is a slightly-edited version of a chapter in one of our books; in this case Net Promoter – Implement the System All of our books are available in paperback and Kindle formats from Amazon stores worldwide, and from your better book retailers.

What about the customers v2

NPS (35) – Sustained Leadership Commitment – Do you have it? – 35th article in a series on the Net Promoter System®

/ , , , / , , , , /

Welcome to the 35th article in my series on the Net Promoter Score and System.

You can read all of the prior articles in the NPS series on our blog page here.

The questions addressed here are about the true level of commitment of your leadership to improving customer loyalty and the associated mid-term financial performance. We will cover how to tell whether you have the commitment already, and what you should do if not.

Are your leaders already committed?

Don’t start by making assumptions. Do your research. If you work for a public company, download and read your annual report. Read your latest earnings announcement. If your company does a quarterly earnings conference call with financial and industry analysts, try to join the call. Download the slides used for the call and inspect them. Assuming your company strategy is stated somewhere in all this material, what is it? Are customers mentioned among the top few priorities? While it could have changed by the time you read this, many consider Apple to be customer-centric, but Apple does not mention customer experience in any of these documents at the time of writing. They are in fact a product-centric company. This does not mean they don’t care about customers at all; far from it. In a product-centric company, work on customer experience simply has to be positioned as supporting product growth.

Prompted versus unprompted support

Reading your CEO and senior leaders’ all-employee emails is close to reading what psychologists call an unprompted response. Unprompted reactions to questions tell you what a person really feels. The prompted reactions of a senior executive bring political correctness into play. Think about what would happen were you to ask these questions in the Q&A section, after your CEO has spoken to all employees. Accept for a moment that all three start with “You have told us your top five priorities, but you have not mentioned…”

  • “…that only 15% of the people managers in the company are women while we represent 40% of the workforce. Do you not think this is important?”
  • “…that most other companies around here give paid time off for employees to do volunteer work in the community. How much do local communities matter to our company?
  • “…our customers as a priority. Where are customers on your list?”

Faced with such prompting, the CEO is forced to reply that he/she is launching a major initiative to double the number of female managers, that we can look forward to an announcement about volunteering, and that customers are central to everything we do, so should not be mentioned separately, or something similar. None of these responses change the reality of the top priorities: the list the CEO gives without prompting. If customer experience is not on your company’s formal priority short list, you can change it over time, with patience and persistence.

Leaders’ job history may work against you

If your company hires leaders from outside, it can pose particularly difficult problems. After all, the executive who has been poached is naturally convinced that he or she has been hired for their beliefs, knowledge and results, as well as the business practices and results of the former employer. If her former company had better results than yours, and did not use the Net Promoter System, it is unlikely that she can easily be persuaded to adopt it. She is more likely to propose that NPS be abandoned in favor of whatever she is used to. This is where your CFO should help, assuming you have jointly proven the financial effect of NPS trends for your company. (This won’t work if the new executive is the CFO, of course.) If the executive’s former company is one of your competitors, NPS should be positioned as a way you are going to catch up and overtake them. That position is hard to argue with, as most executives understand that you can never overtake a competitor by copying exactly what they have been doing.

Do your leaders spend much time with customers?

According to Bain research shown below, the larger the company, the more time leaders spend on internal meetings, rather than directly with customers. They know this and they hate it.

Ask your senior managers how much of their time they spend with customers and how much they would like to spend. If they currently devote more than the proportions below to customers, you can consider them to be committed. All of this has a perverse benefit for the customer experience leader in a large company. If you can come up with simple proposals that let leaders spend more time with customers, and with minimal effort on their part, you will win.

Check your Intranet

Now you know what is being said publicly. Are the public statements about customers backed up by the internal priority list? What has the CEO said in his or her latest email message to all employees? Assuming you have an intranet, go to the home page of each member of the corporate leadership team. Look for their lists of priorities for their teams. Are customers mentioned? If you work for a large company, look one layer down. What do you see? I suggest making a table with all the priority lists and discussing it with your leadership team.

Do the actions match the words?

Now you know what your leaders are saying. Time to inspect what they are doing. Start internally. Does your company hold events for sales teams? Do customers or partners attend and speak at these events? If you have effective monopolies in some product areas, what has been your approach to pricing? Do you put prices up each and every year, as some notorious pharmaceutical companies have been seen to do? If you are in software and perform license audits on customers, is it done by a helpful operations person or an aggressive lawyer? What is your supply chain team’s attitude to their operational metrics like level of fill and delivery to customer commit dates? Is there evidence that both the standard and the performance have been steadily improving over time, or does cost reduction seem to compromise that?

Price gouging — a particularly insidious move

When I first wrote on this subject in 2016, the New York Attorney General had launched an investigation into Mylan Pharmaceuticals, the maker of EpiPens. An EpiPen is used to provide life-saving injections to people in anaphylactic shock, due to allergies. The reason for the investigation is that Mylan have had a monopoly on the product and have increased its price in the USA by 471% since acquiring it in 2007. The CEO’s initial public attitude to this seems to be “Hey… I am just running a business here.”

Daniel Kahneman has written about price fairness and the consequences of unfair pricing in Thinking, Fast and Slow. He gives examples of a hardware store that puts up the prices of snow shovels by a third during a storm, and of an employer putting down the salary of an existing employee due to competition. Very few people consider these practices to be fair. So what? So… he goes on to cover the revenge customers are prepared to take when they see an unfair practice. Indeed, not only will customers take revenge, many other people who are not customers will happily join in to punish the unfair practice and they will all feel great about it. This is what is currently happening with Mylan. Go to their website at www.mylan.com. It purports to show how caring they are. The actions do not seem to match the words.

Corporate goals can conflict with each other

In larger companies, each corporate function can have goals that seem perfectly sensible in isolation, but conflict with each other. I remember one of these situations arising when I was in the HP office in the center of Prague. A bank across the road had a guaranteed 30-minute intervention time for their systems. We held spare parts in our building for them. The local team were unpleasantly surprised by the results of the corporate Real Estate team’s cost reduction goals. We had to leave the expensive office downtown and move to a cheaper location, making it impossible to honor the service time guarantee with existing processes.

A similar exercise led to a proposal to close a remote parts stocking location in the east of Russia. The next-closest office was over a thousand miles away. Our response time guarantee to a local oil company was at risk. Unfortunately, we had to find more expensive solutions for servicing the customers while still allowing the Real Estate team to make their goals. If you work at a company that needs to reduce costs, try hard to ensure that each cost goal that could impact a customer is accompanied by an appropriate customer-centric operational metric.

 

Next time

The next article will be the first in a set of three about securing and sustaining leadership commitment for your work.

 

As is often the case, the above is a slightly-edited version of a chapter in one of our books; in this case Net Promoter – Implement the System All of our books are available in paperback and Kindle formats from Amazon stores worldwide, and from your better book retailers.

Acme-800-square

NPS (34) – Employee development opportunities – 34th article in a series on the Net Promoter System®

/ , , , / , , , , /

Welcome to the 34th article in my series on the Net Promoter Score and System.

You can read all of the prior articles in the NPS series on our blog page here.

Many companies have employee training and development processes. Here is how the Net Promoter System should be included in three common types of company training program. And I have added the latest information about recognized NPS professional certification.

Graduate development programs

Most large companies have formal programs to hire and develop graduates from targeted business schools. Company representatives have stands at university recruitment days. They also proactively contact students in particular classes, trying to position the company as an attractive place to work. Once of the common graduate benefits is a formal training program for new hires, along with guaranteed minimum salary increases for the first year or two. One the one hand, these new hires may rise quickly in the organization. On the other hand, they are relatively unlikely to stay for more than a couple of years.

The customer experience leader needs to contact the leader of the graduate program and supply the NPS training content without being asked. If there is a welcome pack or website for new people, your customer listening and improvement process needs to be part of it. Be open about what customers want the company to improve. I don’t see a reason to directly involve new hires in the improvement work. They simply need to understand what the overall process looks like.

Top Talent programs, Fast Track programs

HR evaluation processes are used to identify ‘top talent’ in many companies. These are the people who have high formal performance ratings and who have been subjectively identified as having strong growth potential. (If you think you work for a company that has abandoned formal performance reviews, you are kidding yourself. They still exist, and have simply been given a different name. There is always some fact-based process for distributing money from the annual salary increase and bonus pools.)

I have been involved in several such programs, and it has always surprised me how difficult the organizers find it to develop the curriculum, when it is done internally. This is a great opportunity. I have found them receptive to proposals to deliver customer-centric content myself. The most common proposal I have made in recent years is to start by explaining the Net Promoter System and the latest competitive benchmark results. The participants are then invited to brainstorm and develop improvement initiatives that they then manage while they participate in the program. The last time I did this at HP, I was given a full day of the three-day opening session.

Individual development plans

Most medium and large companies require managers to agree job plans and development plans with each of their employees. It may be counterproductive to require all managers to include a customer-centric objective, as most will be meaningless. A better approach is to write to managers who are about to start working on the development plans. Provide them with a list of the improvements customers and partners are requesting and suggest which ones might be appropriate for a specific manager’s people. Offer to support the formal documentation and implementation of relevant projects.

NPS professional certification

There are not many sources of professional certification for the Net Promoter System. Bain provides customized training for individual clients, though no recognized certification. Satmetrix has been acquired by NICE and is the only official source of certification for your NPS professionals. Official means recognized as such by the co-inventors, Fred Reichheld, Bain and NICE-Satmetrix. The current certification format is called the Net Promoter Masterclass. The development and implementation of the platform has been managed by former Satmetrix CEO Richard Owen and others from his Owen CX team. Their main certification offer costs $3,000 and includes the online certification test, 12 months access to a large library of material, and access to a forum moderated by Owen CX consultants.

Medallia and some other companies also provide non-official NPS training. Medallia’s it concentrates on aspects that support the use of Medallia software. Not quite ideal.

Conclusion

Inclusion of the Net Promoter System in company hiring and development programs will not happen spontaneously. The customer experience leader needs to find out what development programs exist and get involved at the appropriate time. Some of the timing is predictable, such as when graduates are likely to join, and when HR requires all development plans to be submitted in Workday, or whatever management tool your company uses. Be proactive.

 

Next time

The next article will be the first in a set of three about securing and sustaining leadership commitment for your work.

 

As is often the case, the above is a slightly-edited version of a chapter in one of our books; in this case Net Promoter – Implement the System All of our books are available in paperback and Kindle formats from Amazon stores worldwide, and from your better book retailers.

Ask employees what customers want

NPS (33) – Why not ask your employees what your customers want? (ceNPS) – 33rd article in a series on the Net Promoter System®

/ , , , / , , , , , /

Welcome to the 33rd article in my series on the Net Promoter Score and System. I want to introduce a new concept this time. I call it ceNPS and here is what it is about: Your employees have quite a bit of knowledge about your customers. Why not ask them what your customers want you to improve? Employee ideas tend to be easier to implement and the whole process costs less and is faster than going systematically out to customers.

You can read all of the prior articles in the NPS series on our blog page here.

We have talked about asking customers what to improve, and improving it. We have mentioned the same approach for employees. I now want to introduce a hybrid concept: ask employees what to improve for customers. Let’s call it Customer-Employee NPS, or ceNPS.

Why should the customer be the only source of customer knowledge?

Customers are the best possible source for information about what they would like to see improved. As to how to improve it, well, they leave that up to you. Your employees have a different perspective. Most are aware of things that could be made better for customers. Many also have precise ideas about exactly how to do so. I therefore suggest testing the following survey with your employees to see what value it provides:

  1. How likely are our customers to recommend our company?
  2. Why?
  3. What should we improve for customers and how?

Use the same techniques as for customer research to prioritize the input. I don’t see why this would not work in all industries, bearing in mind that ‘customer’ means the direct customer of the company, which could be a retailer that sells onwards to end customers.

Compare with customer feedback

In a rational universe, there will be at least some overlap between customer feedback and employee views on what customers need. Compare the top five suggestions on both lists. Items that come up high on both lists are ideal; the customer’s expression of the need will match employee suggestions on how to implement the improvement.

Employees will have different ideas

Employees may have insights in areas that would not naturally occur to customers. One example that springs to mind happened shortly after HP set up a centralized pre-sales group in Bucharest, Romania. When the new operation started up, it would be fair to say that the people were quite passive, just doing what was asked. Sales people and resellers asked them to provide price quotations to customers, and they did that work well. Then the team supporting the Benelux noticed something. As negotiation cycles continued, sales people removed things from orders to meet a customer price point. Notably, offers for servers and storage that started with round-the-clock support services were often reduced to a cheaper business-hours-only proposal.

The Benelux team in Bucharest decided to systematically give the customers two proposals: one exactly as requested, and one with the service levels they believed the customers really needed, at a higher price. The initiative was a great success, with about half the customers taking the more appropriate service proposal at the higher price. For customers who accepted, it completely eliminated the situations where they would call during a weekend to be told that their contract did not cover the repair. A customer would never have thought of this double-proposal idea.

Employees should lead the implementation of their own ideas

Unless you have an excellent reason for not doing so, the person who made the suggestion should always lead implementation. Letting anyone else lead eliminates a career development opportunity and allows one aspect of human nature to creep into the picture: rejection of other people’s ideas. Just as the golden rule of project planning is “Those who implement must do the planning”, the golden rule of ideas is that the idea generator must drive the implementation. Yes, it is true that really creative people are usually awful at implementation, and that will be their career development opportunity.

Include your subcontractors and partners

If you use resellers and subcontractors, they should be included in your improvement culture too. Simply follow the same approach as for employees. I know of no example of any company doing this and am sure the reaction will be positive. It is to be expected that your partners will have privileged relationships with some of your end customers. You need to learn from them.

 

Next time

The next article will be about employee development in the context of customer experience work.

 

As is often the case, the above is a slightly-edited version of a chapter in one of our books; in this case Net Promoter – Implement the System All of our books are available in paperback and Kindle formats from Amazon stores worldwide, and from your better book retailers.

So Happy Here

NPS (32) – Employee Net Promoter Score (eNPS) – 32nd article in a series on the Net Promoter System®

/ , , , / , , , , , /

Welcome to the 32nd article in my series on the Net Promoter Score and System. This time the subject is a variation on the Net Promoter Score: the Employee Net Promoter Score or NPS. Bain revised the initial concept just under two years ago so some of what follows may be new to you.

You can read all of the prior articles in the NPS series on our blog page here.

What is eNPS?

Most medium and large companies, NGOs and public sector organizations measure employee happiness in a structured way, on a regular basis. The purpose of such research is usually to help executives to understand what the general employee population would like them to improve. The CEO also gets to find out what improvements the executives would like to see. Almost everyone uses excessively complex survey methods that take too long and provide limited information about improvement suggestions. Applying the Net Promoter System solves these problems.

Darci Darnell of Bain announced Net Promoter for People in a Bain Net Promoter System podcast on November 30th, 2017. It enhances the original eNPS by adding an additional question. Employee Net Promoter Scores used to simply be the answers to the question: “How likely are you to recommend our company as a place to work?” The additional question is “To what extent do you agree with the following statement: I am inspired by my job.” It is measured on a five-point scale.

Bain research showed that verbatim comments relating to recommendation were mainly about basic needs and things about which employees are dissatisfied. The inspiration question produces verbatim responses that are about their higher-level need and suggestions for a positive future. The two rating questions are followed by the standard “Why?” and “What could we do better?” questions. Note that I continue to use ‘eNPS’ to refer to the new format. And it is also the case that the Bain website still (as of September 2019) contains references to the old eNPS format.

How not to measure employee satisfaction

Most companies that measure employee satisfaction do so once a year. Data collection is a significant exercise and a typical survey has about 80 questions. Analysis is often done by a third party. Results are usually available four to eight weeks after the survey period ends. Companies may further delay publication for various real or imagined reasons, such as not wanting them to compete for attention with quarterly earnings messages. Sub-scores are calculated for areas like employee engagement, or compensation & benefits. There may or may not be open questions with free text input.

To use an old saying, the result of such surveys is far too much data and far too little information. Let’s suppose your employee engagement questions score lower than last year, and lower than the average for your industry. Since you only have numbers, you don’t know why it happened. The usual first phase is one of denial, “Well, we had a major management change right in the middle of the survey period, so the change is understandable. Let’s wait until next year to see if we need to act.” If you can get through that sort of reaction, the next phase is one of focus groups. You may have collected data from 10,000 people, but the decisions about what to do are given to a group of ten people, for example. You could just have started there, skipping the survey, if you think that is a good approach.

Benchmarks

There are no recognized benchmark suppliers for eNPS. The Mayflower group in the USA and the RACER group in Germany have set up agreements among member companies to share employee survey data. The IT Survey Group (ITSG) is a recognized source of non-eNPS employee survey benchmark data for the IT industry. Member companies agree to ask 60 survey questions in exactly the same format, and using the same rating scale. They then use the average ratings as a norm for their own scores. There are no verbatim questions in the ITSG data set, so there is no information on what the average IT industry employee would like his or her company to improve.

The 24 companies shown below are the ITSG members as of September 2019. They all use the long-format surveys ITSG membership requires. At HP, we took the comparison with ITSG numbers seriously. We just did not directly know what to do.

 

Members of the IT Survey Group

Let’s make it a lot simpler

The new four-question eNPS format used for Net Promoter for People presents an obvious way of doing simpler surveys. You already have all the demographic information you need about them, so analysis should be easy. Yes, the demographics matter a lot. Men and women tend to give radically different answers, for example. Happiness also tends to decrease with increasing length of service.

No matter what the demographic group, the trend is more important than the absolute number. This brings us to another advantage of the eNPS method. I recommend sampling one-sixth of your population every month, sending new survey requests each and every day. The great advantage this brings is that you will be able to see the effect of different things happening in your company, such as earnings announcements, new product announcements, executive arrivals or departures, and even external factors such as a stock market crash or an election.

Follow the NPS framework

The NPS framework works for eNPS as well. Employee feedback is usually anonymous. Typically, a manager will only receive a report if at least eight of her employees have provided feedback. The Inner Loop then takes place at the level of each team. The Outer Loop addresses company-wide improvements. Regular Huddles are the correct way to drive progress and ensure interlock between inner and outer loops.

Indirect advantage

On top of simplicity and the ability to do continuous monitoring, there is an indirect benefit. All employees will become familiar with NPS. They will easily understand that you are asking customers and employees similar questions and follow the same improvement process. Of course, if you take no visible action on the employee feedback, it will be harder to persuade employees that you are indeed acting on customer feedback. You need to do both.

Cultural differences matter a lot in traditional employee surveys and less in eNPS

While cultural differences matter somewhat in customer surveys, they seem to matter a lot more in employee surveys. CEB (www.cebglobal.com) has published the graph below that emphasizes the point. In short, unrealistic self-perceptions of performance are common. These differences are greater than those found between countries for customer surveys.

Personally, I find the table surprising. North Americans use ‘curve grading’ in their schools. Curve grading means that a fixed proportion of a class gets an ‘A’ grade, a ‘B’ and so on, no matter what the objective level of the class is. Most Americans therefore grow up having a realistic view of how they compare to others in the classroom. Apparently, this knowledge is discarded when they enter the world of work. Curve grading is rare in the other countries on the list. Could these facts be related? I could speculate, but don’t really see how they could be.

In short, unrealistic self-perceptions of performance are common. These differences are greater than those found between countries for customer surveys.

CEB employee self rating by country

An additional communication challenge

We have seen that you should not compare countries to each other in your reporting. A fair way of reporting on countries would be to show only the trends, not the absolute numbers. Just leave the numbers off the y-axis in your graphs. There is an additional communication challenge with eNPS. I have generally seen eNPS scores that are about 20 points lower than the customer scores for a given business. What that means is that eNPS scores are far more likely to be negative numbers than are customer scores.

As previously advised, if your first report on a topic only contains bright red negative numbers, your audience will intuitively deny the reality of your report and ignore it. Rather than perverting the system by (say) adding 20 points to all the numbers to ‘normalize’ them, I suggest devoting the first report entirely to the improvement suggestions and the answers to the Why question. You can supply the positive or negative eNPS trend numbers from the second report on, without necessarily providing the absolute scores at all. You may like to believe your audience will have a rational response to a set of negative numbers, but they will not. Intuition rules. Deal with it.

Conclusion

Use eNPS as your primary ongoing survey methodology and execute it continuously. If you (mistakenly, I believe) think you get value from participating in a benchmark consortium, continue to do that huge survey for a year or two, comparing the actions you take from each survey type. I am confident you will discard the consortium survey, particularly as it does not give you direct insight on the performance of any individual competitor. If you genuinely want to improve things for your employees, just ask them what to do, then do it. The Employee Net Promoter System and Score are the most efficient and effective way of doing so.

 

Next time

The next article will be about a new concept: asking your employees to tell you what improvements customers would like to see.

 

As is often the case, the above is a slightly-edited version of a chapter in one of our books; in this case Net Promoter – Implement the System All of our books are available in paperback and Kindle formats from Amazon stores worldwide, and from your better book retailers.

So Happy Here

NPS (31) – Does employee happiness matter? – 31st article in a series on the Net Promoter System®

/ , , , , / , , , , /

Welcome to the 31st article in my series on the Net Promoter Score and System. This time I dig into the question of the relationship between employee and customer satisfaction. The answer is not what most people believe. While this series is about the Net Promoter Score, I have used American Customer Satisfaction Index data for the analysis as I was able to study 398 companies. I have since been able to do the same type of study using Temkin NPS benchmarks for 223 companies, with very similar results. However, I still consider the ACSI-based study to be superior, just because of the number of companies and industries I was able to include.

You can read all of the prior articles in the NPS series on our blog page here.

Contrary to popular wisdom, happy employees are not terribly important to customers. My research shows that employee happiness contributes about 6% of customer happiness. This does vary by industry.

Source of data

In February 2019, I compared American Customer Satisfaction Index data with Glassdoor.com employee ratings for 398 large companies that sell to US consumers. The research has been described in detail in our companion book Customer Experience Strategy – Design and Implementation. Using linear regression, just 6.2% of the variability in customer satisfaction is due to employee satisfaction. For the stats experts, the number does not change with quadratic regression. This was our third annual study on the subject and prior results were similar.

Reasons for the surprise

There are a number of reasons the link between end customer and employee happiness is not as high as you might expect:

  • First and foremost, a happy employee is not necessarily an engaged employee. A person can be happy with their short commute to work, their pay, and the free food in the company restaurant. None of these involves doing anything positive for customers. Engagement is a more complex concept, with no standard definition. However, every study I have seen shows a positive relationship between employee engagement and a range of things, such as financial results and customer satisfaction.
  • Many brands do not actually interact directly with customers at all. This is the case for franchise businesses. The company grants a right to use a brand in a geographical area to a franchise operator and the franchisee is the one who interacts with the end customer. Even the largest corporations that deliver physical goods do this to an extent. Almost none cover every country in the world themselves, depending on local third-parties to cover their gaps. The happiness of the employees who have no interaction with the end customer is of course invisible to them and has no impact on their NPS scores received from those end customers. The parent companies should be more concerned with the happiness of franchisees. There is no easily-available public data on franchise operator NPS.
  • There are web-only brands where you are unlikely to come into contact with a human. Amazon is a company with a great ASCI score of 82, and a relatively ordinary Glassdoor score of 3.8.
  • Some brands have monopolies and can afford to treat customers poorly. This seems notably to be the case for many cable TV companies and Internet Service Providers. Their customer and employee satisfaction scores seem to be opposites. Who would have thought it possible? Happy employees and unhappy customers. I suppose another explanation is that these companies provide services that we only notice when they are not working. When everything is OK, we don’t think about them. When our internet access is down, we quickly get angry.

There may be sectors where it matters a lot

My research suggests there are high-touch industries where employee engagement makes a difference to customers. There were 17 supermarket chains in my study, and the variations in employee happiness explained 49.1% of the variations in ACSI scores, (as measured by the r-squared number.) For hotels, the number was highest of all, at 68.1%. However, those companies include hotels operated by franchisees, rather than employees of the parent company, and ten is not a great population for calculating correlation coefficients. Perhaps surprisingly, the number was just 7.2% for specialty retail stores. There were 25 companies in this category. I have to speculate that this may be due to the negative environment created by competition with eCommerce. Bearing all of these factors this in mind, I nevertheless selected 129 of the 398 companies covered as ‘High touch’ and found that variations in employee satisfaction explain 8.8% of the variations in customer satisfaction, using linear regression, and 16.8% using quadratic regression. Still quite low really.

Think about it

If you were to assemble a team of people and ask them to brainstorm potential factors that affect customer happiness and loyalty, I believe you would come up with a long list. Employee satisfaction would probably make the list, but I doubt it would be in the top five. Things like the reliability of your products would be higher up on the list. It is only when the question is asked in isolation that people say employee satisfaction is the most critical factor. It is a case of an intuitive reaction, rather than a rational one. It would be tempting to believe that employee satisfaction is a hygiene factor, meaning that unhappy employees do indeed have a negative effect on customers, and that there is a diminishing rate or return on improving happiness, making no difference at all above a certain point. However, the ACSI analysis does not support that theory either. The relationship is close to linear at the cross-industry level.

Full details

If you would like to read about this in far more detail the full blog posts, which are slightly-edited versions of chapters in our book Customer Experience Strategy – Design and Implementation are here and here.

 

Next time

The next article will be about one way of measuring employee satisfaction_ eNPS.

 

As is often the case, the above is a slightly-edited version of a chapter in one of our books; in this case Net Promoter – Implement the System All of our books are available in paperback and Kindle formats from Amazon stores worldwide, and from your better book retailers.

Text analytics

NPS (30) – Evaluating text analytics software – 30th article in a series on the Net Promoter System®

/ , , / , , , , , , /

Welcome to the 30th article in my series on the Net Promoter Score and System. This is by far the longest and deepest of the set. It includes an example of a real-world analysis and selection done a couple of years ago. The focus was free software, including software that had free trial versions. My daughter (who also designed the covers of the first three books) did the project as part of her bachelor’s degree program. The important part of the project is the selection principles she used. Software has moved on somewhat since then and at least one of the programs she compared no longer exists. Nevertheless, I believe the selection criteria are still valid today and can be used just as well for commercial software. And of course if any software provider mentioned below would like to provide an update, that would be most welcome.

You can read all of the prior articles in the NPS series on our blog page here.

Introduction

The great advantage of the NPS feedback format is that it is short, with a rating question and one or two other open questions. Like most survey types, the purpose is to find out what customers care about and what you need to improve. The main challenge of the open text response format is the need to analyze what people have written, ideally without any human bias. I would like to cover some general principles about using text analytics software, then present a real-world case where such software was analyzed as part of a business school degree course. The overall intent is to provide you a way of comparing and selecting software for your own use. There is no single universally-correct answer. Your answer will depend on factors like volumes, languages and budgets, among others.

General principles about text analysis software

Here is what I believe text mining and analysis software must be able to do, with increasing levels of difficulty as shown below the bullet list. Software should be graded in each area:

  • Accept inputs in a variety of formats such as Excel, SurveyMonkey, SalesForce, .csv, video and audio files and so on. The more the better.
  • Recognize inputs in a variety of languages. Again, the more the better.
  • Identify concepts in a meaningful way. There are four levels:
    • The first is simple word frequency, with no filtering. Words like ‘the’ or ‘very’ could come up.
    • The second recognizes nouns and discards articles and conjunctions. ‘Heart’ and ‘Attack’ would be separate topics at this level.
    • The third level is multi-word topics, such as ‘Heart Attack’ or ‘Cardiac Arrest’.
    • Finally, words that mean the same thing need to be grouped into a single concept. For example, “support website” and “support webpage.” This avoids long lists of concepts.
  • Identify sentiment to determine whether respondents refer to a concept in a positive, negative or neutral way. Sentiment analysis determines which attributes a customer associates with the concept.
  • Identify themes by combining concepts with sentiment and/or linguistic analysis. ‘Fatal heart attack’ or ‘Disgusting undercooked food’ might be themes. ‘Bad food’ is a different theme to ‘good food’. This last example illustrates why simple word frequency software is not useful for survey analytics. Just getting the word ‘food’ a lot will not tell a restaurant owner much.
  • Group similar themes to avoid excessive granularity. ‘Serious heart attack’ and ‘Major cardiac arrest’ are grouped together in this step. It may involve use of industry-specific libraries or taxonomies. Some software is delivered with such libraries or lets you add your own. Other programs use data-driven approaches to determine similar words based on how your customers use them in a sentence. Google, for example, uses data-driven Natural Language Processing to make sense of your search queries.
  • My experimentation so far has shown that correct handling of pronouns turns out to be very difficult indeed for software, remaining easy for humans. More on the problem later.

Text analysis for surveys

 

State of play in early 2017 – The principles still apply

What follows is an example of an actual comparison and evaluation of analytic software. The software evaluated was for a specific need and had to meet specific criteria. Your needs and criteria will be different. It does get somewhat technical. You may like to just skim over it, or jump straight to the results at the end of the article.

My daughter Claire did a business school project about text analytics for Net Promoter Score surveys in mid-2016. She got permission to survey all business school students, then used a wide variety of software to analyze the text responses, comparing them with human analysis. The survey was a classical NPS survey. She used SurveyMonkey to send it, and there were three questions, plus demographics.

The questions were “How likely are you to recommend the business school to someone choosing their third-level studies?” “Why?” and “What should the business school improve?” She resisted attempts by others to add rating questions, and the survey results validated this resistance. One of the top suggestions was something that would never have come up using a closed survey.

Overall objective of any survey analysis process

Unless you are only interested in tracking a metric, for example to reward people, the fundamental purpose of any survey analysis is to find out what your top improvement priorities should be. The winner of any comparison should therefore be the process or tool that answers that question. The advantage of software over humans is that the only bias is that caused by the programmers. I admit that such bias can be considerable, particularly when selecting the algorithms to be used to categorize text.

A hard historic lesson on decision criteria

In all scientific experiments, it is critical to set up the methodology before you have any data. This was made very evident by changes in US rules on drug approvals. Experimenters are now required to provide details of the exact methodology and data analysis process before they start clinical trials. The results have been dramatic. If you go back 40 years to when researchers could decide how to run the analysis after they had the data, over half the trials were successful. Now the success rate is just under 10% , as people cannot remove data or change their analysis after the fact because the results do not suit them.

The seven-country study farce that affects our lives today

The most famous study that suffered from post-hoc adjustments was the legendary and highly influential ‘Seven countries study’ by Ancel Keys on the relationship between diet and heart disease by country. The study was done in 1958 and continued for 50 years. It showed a directly linear relationship between saturated fat consumption and heart disease. In 2009, Robert Lustig found that Keys had cherry-picked the seven countries from a data set that included either 21 or 22 countries. He apparently discarded the data from the 15 countries that did not support his theory. A separate piece of work also showed that Keys’ results could be better explained by differences in sugar consumption, which Keys had not studied. There is a nasty history of scientists who took positions opposing Keys losing their research funding and even their jobs.

Decision criteria for text mining and analytic software

The people at Sift, now called Keatext , produced a nice eBook on how to select this kind of software. There is a danger of bias here, in that their own software naturally meets their own criteria. For her study, Claire decided to use some of the Sift criteria, deciding that there were a few factors that did not matter much. The result of her work was these decision criteria:

  • The ability to access, read and extract words in common file formats and in multiple languages.
  • Show how often topics occur.
  • Ability to attach sentiment to topics in a useful way. Useful means that it is easy to understand what the sentiment refers to.
  • Ability to group similar topics together, even if the words are somewhat different. This is what allows you to summarize the actions that need to be taken from a survey.
  • Ability to refine the selections for reporting purposes.
  • Nice graphics and other features that make presenting the results easy.

Each item was scored on a 1 to 6 scale with 6 being the maximum. What can I say? The Swiss education system grades this way, so the scale was natural for her.

Scope

All software analyzed had to be either free or available in a fully featured trial version, and all had to at least claim to be up to at least some of the task. This led to some notable exclusions, especially commercially available text analysis solutions which we discuss in the next chapter. The trial version of Watson does not do text analytics. She was unable to find NPS survey analytics examples for IBM’s BlueMix software on Github or the IBM site. Here is the list of solutions that are described in more detail below:

  • Humans
  • Data Cracker
  • WordyUP
  • Wordclouds.com
  • Keatext
  • SurveyTagger
  • Etuma Feedback Categorizer
  • SPSS (from IBM)
  • Lexalytics
  • MeaningCloud (Excel Plug-In)
  • Text2Data (Excel Plug-In)
  • Haven OnDemand

Humans

With all their defects, the baseline for all comparisons was human beings. Some students were asked to scan the survey responses and group the ones they felt to be the most important. Human bias emerged quickly. One of the students looking through the improvement suggestions was French. Several people (presumably trying to be funny) had suggested that fewer French students should be allowed to attend the school. The French student going through the responses insisted that this was the most common suggestion. It was not, and did not appear on the lists given by other students.

Nonetheless, the students were reasonably consistent. Among the reasons for ratings, humans picked the quality of the education as the top reason, followed by the beauty of the campus. Some software agreed. The consensus of the human analysis was that the top suggestion by Promoters, Passives and Detractors was to announce the exam dates earlier. No software available without charge identified this as the top improvement suggestion. Survey responses were mainly in French, with some in English and just one in German. The humans understood all three without difficulty. Priorities for the why and improve questions are shown for the first four humans.

 

 

Humans don’t scale well

This is one area where the results of the business-school survey may differ from more common needs. The small number of responses meant it was realistic to ask humans to do the analysis. If your needs involve tens of thousands of lengthy responses, using people to do the analysis will be difficult. Humans can review about 15,000 words an hour, and the quality of the work goes down as they get tired.

Main challenge when humans help software

Generally, tagging responses to manually categorize them introduces human bias and is not a good idea. In one part of HP, tagging support survey responses worked well until a second person was needed to handle the volume. Since the two did not agree on the categories, trend data was disrupted. Adding a third person made it worse. You can make manual tagging work if you invest in training.

Here is what humans can do that no software can as yet

Humans can handle pronouns and understand context. Let’s consider pronouns first. No software tested could handle pronouns correctly. Take this unrelated example. “The cat walked into the room. It was empty.” No software tested could identify what the ’it’ was. While it might seem simple to write code to force the software to revert to the prior noun whenever it encounters a pronoun, there are many exceptions to this rule. This brings us to the second thing that does not present any difficulty for humans. Take this example. “The cat walked into the room. It was fluffy.” Humans know that fluffy almost certainly refers to the cat rather than the room.

Humans are good at recognizing and adjusting for context. Both humans and software would be confused by “The cat walked into the room. It was dirty.” Good grammar suggests that the room was dirty, but it is difficult to be certain. I suppose this is one of the challenges that Natural Language Processing software will need to address in the future.

DataCracker

DataCracker is available at www.datacracker.com. It is free for surveys with up to 100 responses. It works by looking at individual words. It is not able to identify multi-word topics or concepts. It is probably the software that is able to import data from the widest variety of formats. The business-school survey had over 250 responses, so the full set of answers could not be loaded into the free version of the software. You can see a screenshot of the main text analysis view below. It is a typical word cloud, with font size indicating relative frequency. Since it does not understand language, irrelevant words like “la”, “les” “très”, “et” and so on come out frequently. (The, the, very, and). As with all word clouds, you can see in Exhibit 9.3 that “interesting” is a common word, without any indication of what exactly is interesting. Unlike WordyUp and Wordclouds.com, you can ignore words by clicking them on the list on the right. This of course allows you to bias the results in any way you like.

WordyUp

WordyUp goes beyond the basics of word clouds by showing the relationships between different words. However, the free version at wordyup.com is limited to 500 words. The central challenge the software faces is the same as for many word clouds: irrelevant words like ‘because’ take priority if they occur frequently, as shown below. At least in the free version, there is no way to exclude any particular word. The only form of data entry is to copy text into the appropriate box on their website. The sample below is from 500 words that were responses to the “Why?” question in the survey. The grouping of words about the campus is well done. Groupings for the improvement question were far less useful.

 

Wordclouds.com

For pure simple and free artistic word clouds, it is hard to beat wordclouds.com though Wordle produces similar results. Wordle and Wordclouds.com are so similar that it was not useful to include both in the analysis. The software works on any device that can display HTML. The output can be displayed in a variety of attractive shapes. It treats ‘Good’ and ‘good’ as two different words. Words can be excluded manually. You need to exercise caution when using the control slider to resize the cloud, as differing words are excluded from the cloud depending on the size of the graphic you choose to display. Some of the most popular words are excluded in some formats, and there is no apparent reason behind the choices. Words can be excluded or added manually, and the weight of each word can be modified. Exhibit 9.5 shows the output using an English translation of the “Why?” responses.

 

Keatext

Keatext is another cloud-based solution. A fully functional 14-day trial is available at www.keatext.ai. The software does sentiment analysis. It ranks words by frequency and shows the positive, negative and neutral sentiment proportions using an attractive graphic format. Clicking on any of the circles on the right takes you to a diagram that clusters words that have been used together with the word in question. It does limited word grouping. In the screen capture below I suppose the fact that ‘professeur’ was in French prevented it being grouped with the similar words at the bottom of the table.

For NPS surveys, you can select the question you want to analyze, and can filter using demographic fields. Excel was used as the input format for this study. The software only works correctly in English, as it cannot display accented characters, at least using Chrome. Overall, Keatext is a nice improvement on word cloud software. It is Natural Language Processing software and it is constantly being improved. In my opinion, the most important next step would be to improve the ability to extract multi-word topics.

SurveyTagger

SurveyTagger is no longer being developed as Etuma is concentrating on their enterprise solution, Etuma Feedback Categorizer. (See the additional paragraph on the new software below.) The software is an evolution of earlier software used to manually tag survey results. SurveyTagger tags automatically and works in many languages that use western character sets. Since the authors are Finnish, I suppose they realized they needed to get beyond Finnish to be successful. No matter what the input language, the results are displayed in English, as shown below. There are differences in analysis due to translation. As mentioned, we translated the survey results into English for software testing. The diagram below shows the analysis of the “Why?” question when the input was in English and in French. Note that there are at least some multi-word topics, though not many. Actual survey responses can be shown below the graphs. You can drill down a little by clicking on an individual item.

 

Etuma Feedback Categorizer

While it is not immediately obvious that it is possible to try out this new software for free, it can be done by sending your file to Etuma. (The software was tested after the business-school project was completed, so has not been included in the results table.) They have taken what they learned from Survey Tagger and applied it to this more sophisticated solution. The software does what it claims, namely categorizes survey feedback. It adds some useful elements to the categorization.

In the first view in Exhibit 9.8, you can see the Promoter / Passive / Detractor NPS status of people who have mentioned each topic. In all views, the issue with single-word topics remains evident, especially compared with the human analysis shown above. The whole subject of the beauty of the campus on the shores of Lake Geneva does not really show up. ‘Appearance’ is there, but appearance of what?

A nice touch in the Feedback Categorizer is the new ‘Ambience’ view. It provides a sentiment-based graph making it easy to know whether a topic has been mentioned in a positive or negative way.

The same data is used for the word cloud view that is combined with the Ambience bar chart shown above.

 

 

SPSS Text Analysis for Surveys

IBM acquired SPSS a few years ago. SPSS is generally viewed as the premier statistical analysis software. As you might expect, their SPSS Text Analysis for Surveys software is detailed and complete. Perhaps too detailed and too complete for most people, as shown in Exhibit 9.10.

The trial version works in English, is fully functional and can be downloaded from http://www-03.ibm.com/software/products/en/spss-text-analytics-surveys.

SPSS comes with several taxonomies and these are essential to the correct grouping of topics. There was no education taxonomy in the trial version, so the brand image data was used instead. There are lots of options. Their website explains how to drill down into categories and combine them manually. It is possible to export the results into SPSS for statistical analysis and for better graphing capabilities.

Verdict: too detailed for the casual user. Too difficult to get to grips with the software and summarize the actions that need to be taken from a survey.

 

Semantria by Lexalytics

This was the best of three Excel plug-ins at one important task: the identification of multi-word topics. It produces useful multi-word word clouds and works well in French. The word clouds are rather exotic Excel Pivot Charts. As with all Pivot Charts and Pivot Tables, clicking on any element takes you to a new sheet with the corresponding underlying data.

The software performs sentiment analysis and provides a sentiment score. It did a nice job on the “Why?” question and did not do so well with the improvement suggestions. The improvement suggestions varied widely and no software was great at grouping similar multi-word topics together. Humans are much better at this so far.

The screen-captures that are combined in Exhibit 9.11 show the main interface and the clouds for both open-ended questions. The people at Lexalytics have released a version that is specific to survey analysis in late 2016, too late for the study. It currently supports 22 languages.

 

MeaningCloud

MeaningCloud is a free Excel plug-in available at www.meaningcloud.com. The authors suggest using it for surveys and social media analysis. It does some sentiment analysis and works in French, and of course in English. As supplied, it is quite interesting, but does not classify some responses correctly. By that I mean that some of the positive and negative sentiment categorization is opposite to the sentiments actually expressed.

You are encouraged to build up your own taxonomy. It was clear that if you build a taxonomy for your industry, it should do a good job of correctly categorizing responses. It was not worth investing the necessary time to build an education taxonomy for this single survey.

The image below is a screen capture. If you intend to do multiple surveys in a narrow domain, you may find it worthwhile to build up a taxonomy and the software will be more useful.

 

Text2Data

Text2Data is an Excel plug-in that extracts single and multi-word topics and does sentiment analysis. It does not work in French and the screen capture below is from the English translation of the survey responses. The problem with the sentiment analysis here is easy to understand from the screenshot. The sentiment words are shown, but not what they refer to. ‘Outdated chaotic’, OK, but what is outdated and/or chaotic? You have to go back to the individual responses, so the summarization really does not work well. In short, the software was not immediately useful and was not explored in greater depth.

 

Haven OnDemand NPS analytics

Unfortunately, now that we are in September 2019, this application no longer exists. It was written using the Haven OnDemand cloud-based Application Programming Interfaces (APIs) and was free to use. The code for the application has been open-sourced on Github at https://github.com/hpe-idol/netpromoter. Haven OnDemand was developed in the software division of Hewlett Packard Enterprise. HPE announced that the software division would be merged with UK company Micro Focus during 2017 and set up as an independent entity, 51% owned by HPE. Unfortunately Haven OnDemand did not survive the transition. It still exists behind the firewall, but is no longer publicly available.

In English, Haven OnDemand did the best job of any software investigated, though it did not do as well in French. Since it finished second overall in the study, let’s look at it in a bit more detail.

Haven OnDemand background

Haven OnDemand was a set of cloud-based APIs that could be strung together to create powerful applications. Some were very entertaining. If you search YouTube, you should be able to find a Nerf gun that has been programmed to do facial recognition.

The NPS application used only IDOL functions. IDOL stands for Intelligent Data Operating Layer and came from Autonomy Inc., which was acquired by HP. IDOL is used for analyzing all types of unstructured data, including video, audio, and of course text. It works to some extent in over 100 languages, including Vogon and Klingon. (Software programmers do not get out much and tend to like science fiction. I can relate to that, based on overnight stints in the computer room while I was still a student) The sentiment analysis API that is used in the NPS application works well in 14 languages.

A very partial list of the APIs is below, followed by what the output looked like.

 

Distinguishing features

Compared to everything else tested, the Haven OnDemand application did three things best:

  • It grouped related topics together, even if the words used in the topics are not exact matches.
  • It allowed you to exclude single-word topics from your results. Assuming you have sufficient data, this works particularly well in English, not so well in French, Italian and Spanish. The Mandarin tester told me it did an OK job there too, though the concept of a single word is different in Mandarin.
  • The bar chart presentation was the easiest to understand of any software tested.

Like all other software tested, it was poor at pronoun handling.

Results

The illustrations below are the scoring tables my daughter used to evaluate the software tested for her project. Since Etuma Feedback Categorizer was tested later, it is not included in the table. In principle, it would score lower than Survey Tagger for access, since it cannot be used without assistance in a demo version, and higher for filtering and presentation.

Guess what? Humans win! We can’t be replaced (yet). We win because we are still better at understanding how to interpret pronouns and give context to what we read. Note that if your survey needs differ from those of the study, your ranking will differ too.

 

Next time

The next article will cover an example of commercial software that does indeed satisfy all of the criteria mentioned above.

 

As is often the case, the above is a slightly-edited version of a chapter in one of our books; in this case Net Promoter – Implement the System All of our books are available in paperback and Kindle formats from Amazon stores worldwide, and from your better book retailers.

#103 – Warmth and competence – 100 best CX books – Freakonomics – ACSI

/ , , , , / , , , , /

 

Welcome to the 103rd Customer Strategy newsletter. The five topics this week are:

  1. More on the ‘warmth and competence’ model, after recent personal experiences
  2. Our books on list of ‘100 best customer experience books of all time’
  3. Latest blog posts – Deep series on NPS continues
  4. Notable items from other sites – Freakonomics on Trader Joe’s, ACSI on car satisfaction
  5. Looking forward

More on the ‘warmth and competence’ model

I recently had two experiences that reminded me of the ‘warmth and competence’ model of customer experience, as proposed by Chris Malone and Susan T. Fiske in their excellent book The Human Brand. To simplify, our expectations of experiences delivered by humans is that they should be both warm (by smiling, for example) and competent. Here are two recent personal examples of when one of the two factors was missing:

I scraped my car’s wing mirror in a parking lot. I phoned the garage where I bought the car to arrange an appointment. The call was transferred to the receptionist as nobody in the service department was available. (The people in question sit ten meters / 30 feet from each other.) The receptionist passed a message on to the service people using their internal messaging system. The service people had not phoned back 24 hours later so I drove over to arrange an appointment. Note that my wife and I have been buying cars from this garage for the last 25 years.

When I got there to ask for the appointment I was extraordinarily well received by the service people as has always been the case. However, I asked to speak to their manager to discuss the lack of a call-back. He told me that their workload was such that calling back could take up to 48 hours. He suggested I use their online service reservation form to be sure of an immediate response. I went away feeling they were ‘warm’ but not ‘competent’.

That was about six weeks ago. Last week a message came up on the dashboard saying a regular service should be scheduled. I emailed the main service person I had been interacting with to ask for an appointment. After 24 hours without a response I remembered the online service request form. I filled it out. No response 48 hours later and still no answer to my email, so I went over there again to reserve in person. Once again I was warmly received and made the appointment. However, it is clear that a change of general manager of this family-owned business about three years ago has meant cost-cutting and staff insufficiencies that have destroyed the customer experience associated with servicing the cars. Today my wife and I decided that we will never buy another car from them again.

The second experience was simpler and I am sure you have all experienced this sort of thing. A family member lives in Neuchâtel and we sometimes eat at a simple restaurant that is the closest one to where she lives. For the second time we experienced a waitress who was competent but not warm. On our last visit her very first words to us were “Are you ready to order?” No “Hello” or any such niceties. So… I removed that restaurant from the contact list in my phone. No more reservations for them.

The bottom line: we consumers of products and services expect the people who provide them to be both warm and competent. Without both, we leave, if we have a choice. (I recognize that government services and other monopoly situations exist where we have no choice.)

Our books among ‘100 Best Customer Experience Books of All Time’

I discovered this by complete accident on Monday and had no idea that such a list even existed. What can I say? Book Authority publishes lists of top non-fiction books in various categories. They determine their lists using proprietary algorithms that are based on sentiment, sales volumes, recommendations and other factors. The algorithm is not public.

Our first book, Customer Experience Strategy – Design and Implementation, and our third book Customer-Centric Cost Reduction both make the top 100 list. It is possible to click to create a list by estimated sales numbers for the category and both books are also in the top 100. I noticed that Net Promoter – Implement the System has the same rating as the last few books on the list, so must have just missed the cut.

I suspect our particular thanks are due to book readers who have written Amazon reviews and / or publicly recommended them on Twitter or elsewhere. I will be interested to see how the list evolves over time. You can consult it here.

 

Our latest blog posts

The posts on this list are part of my extensive and deep series about the Net Promoter System. Older posts are still available on the blog page.

image NPS (29) – The research / survey infrastructure you need – 29th article in a series on the Net Promoter System®

A simple article that describes the infrastructure you need to send surveys to customers and receive the answers. The simplest solution for PC users who don’t have a lot of customers is Outlook mail merge. Things quickly become more sophisticated…
image NPS (28) – The team(s) and tools you need to be successful – 28th article in a series on the Net Promoter System®

This time I introduce the part of the Net Promoter System framework that is called ‘Robust operational and analytic infrastructure’. It is all about your team and tools…
image NPS (27) – How not to deal with a Detractor (Nike and me) – 27th article in a series on the Net Promoter System®

A person story about how Nike lost me as a customer for one of their non-core product sets. And of course this has made me less loyal to Nike overall…
image NPS (26) – Activating Promoters (so they really do recommend you) – 26th article in a series on the Net Promoter System®

This time the subject is money! We are talking about the Net Promoter System. You have identified people who say that they definitely would recommend you. So how do you stimulate them to actually do so?
image NPS (25) – The Outer Loop and the Huddle – 25th article in a series on the Net Promoter System®

The subjects this time are the Outer Loop and the Huddle. The outer loop is the process for dealing with improvement opportunities that are common to many customers. They can take time and require investments across several functions…
image NPS (24) – The Inner Loop – 24th article in a series on the Net Promoter System®

The subject this time is the Inner Loop; the process for quickly addressing the concerns customers express when you send them feedback requests. In simple cases you need to deal with their issues right away. That is what the Inner Loop is all about…

Notable customer experience items from other sites

Freakonomics – Should America be run by Trader Joe’s?

Yes, I am a huge fan of Levitt and Dubner. Their first book was my introduction to the whole world of behavioral economics. This podcast episode is a re-broadcast of a story I found particularly insightful. Readers based in North America and followers of the American Customer Satisfaction Industry benchmarks will know that Trader Joe’s has the highest customer satisfaction rating of any US-based supermarket chain. Yet it provides far less choice than other supermarkets. Why does this make sense? It is all explained in the podcast. I guarantee you will learn a lot!

Worth a listen here.

ACSI – Customer satisfaction crashes across the automobile industry

The American Customer Satisfaction Industry people just published their 2019 results for all major vendors that sell cars in the USA. Most brands had lower satisfaction than in 2018.

Have a look and their press release to see details here.

 

Looking forward

Book sales continue to do surprisingly well, so thank you. June was the best of the three Junes since the books first appeared. July was the best July. And August was our best August. I have to say that I don’t know why this is happening. While the rather attractive NPS series could be an explanation, our strategy book is slightly outselling our NPS book. Insights welcome.

Here are links to all of our books on Amazon.com. Kindle versions are available in all stores. Print versions are available from the major stores only. And as of two weeks ago, you can find the books, or at least order them in many bookstores. If you have already read any of our books, please write reviews on Amazon.

Customer Experience Strategy – Design & Implementation

Net Promoter – Implement the System

Customer-centric Cost Reduction

“So Happy Here”: The Absurdist but Essential Guide to Better Business (Color edition)

“So Happy Here”: The Absurdist but Essential Guide to Better Business (Black & White edition)

Please share this newsletter with your friends and colleagues and encourage them to sign up for it here. I have put links to past newsletters on the subscription page. Finally, please feel free to change or cancel your subscription using the link below.

You can also email me, Maurice FitzGerald, at mfg@customerstrategy.net.

To change your subscription, click here.

 

Survey infrastructure

NPS (29) – The research / survey infrastructure you need – 29th article in a series on the Net Promoter System®

/ , , / , , , , , /

Welcome to the 29th article in my series on the Net Promoter Score and System. What follows below is my advice about the survey infrastructure you need to be successful. You can read all of the prior articles on our blog page here.

Unless yours is a very small company, you will not be able to implement and operate an effective Net Promoter System without appropriate technology. To the extent possible, you should use Cloud-based (SaaS) solutions that do not require you to invest in hardware. Among other things, an advantage of this approach is that you will not be constrained by a corporate IT budget and can advance on your own. Cloud software is always up-to-date. For the sake of this discussion, Customer Relationship Management software, used by sales and call center people, is out of scope. I will only consider standalone solutions.

Simplest possible research infrastructure

You can use Excel with Outlook mail merge to send surveys, and that can work in small volumes, though the Outlook mail merge function does not exist in the Mac version. Mail merge should work with Open Office too. The primary challenge is ensuring it all looks OK on a phone. Here are the advantages of this simple approach:

  • You probably don’t have to spend any money to get into motion.
  • Assuming small volumes, you can simply ask people to reply to the email to send their survey responses.
  • The emails can go directly from anyone you choose, such as a product manager, or from a process mailbox in the name of the product manager.
  • If you have designed your survey correctly, customers should be able to see the entire survey on a single screen when they open the email. This is almost impossible when using a link to a web-based survey. When customers see the full survey, it improves their belief in your statement that it will only take them a couple of minutes to respond. That in turn will improve response rates compared to a web link.

There are disadvantages, even in small volumes:

  • Managing opt-out lists is difficult. Any survey you send should include a way for the customers to decline the survey and to tell you they no longer want to receive any surveys. If you work from a single Excel contact list, you can probably handle it correctly. If multiple people send surveys from multiple contact lists, it can be difficult to coordinate opt-outs across lists.
  • To do any analysis of the results, you will have to copy-and-paste or otherwise transcribe what has been sent back.

Basic survey infrastructure

Getting beyond the extreme mentioned above, I recommend using SurveyMonkey and MailChimp for basic needs. There are a number of similar alternatives that should work just as well. I just have not tried them all.

SurveyMonkey advantages:

  • There are different levels of monthly SurveyMonkey subscriptions, going from free to advanced solutions with no customer-visible evidence that you are using SurveyMonkey. (Really, they could have chosen a name that would make them sound less like amateurs.)
  • Higher subscription levels include simple text analysis.
  • It runs in many languages.
  • There are ‘certified’ NPS templates.
  • Lots of people use it, so lots of advice is available.
  • You can download results as Excel files for your own analysis.
  • You can send emails directly from their service.
  • If you don’t like it, don’t renew your monthly subscription.

SurveyMonkey disadvantages:

  • If you use a separate email solution, you can only provide a link to the SurveyMonkey survey in your email. There is no way of embedding the rating question in your email, so response rates will be lower (less than half) than with solutions from Medallia, Promoter.io and others that embed the rating question in the outbound message.
  • When I say the text analysis is basic, it really is basic. So is Medallia’s, at the time of writing.
  • Some nice tracking options are only available if you use SurveyMonkey’s outbound email service.

MailChimp advantages:

• MailChimp lets you see who has opened your messages and who has clicked on any link they contain. This is particularly useful for A/B testing. By this I mean, for example, sending email with different subject lines to each half of your mailing list to see which subject lines get better open rates, and doing the same for different requests to take your survey. You should only change a single thing at any time to have a valid A/B test.
• If your list has up to 2,000 customers and you want to send no more than a total of 12,000 emails per month, you can do it for free. The paid plan starts at $20 per month for unlimited emails to 500 customers, for example.

MailChimp disadvantages:

• MailChimp tracks recipients’ actions, but does not seem to disclose that it is doing so, at least by default. I am already aware that this has been raised as an issue with the national data privacy organization in one European Union country.

More advanced solutions

Two capabilities are the most critical when considering advanced solutions for managing NPS surveys. For clarity, I am talking about software that only manages surveys. There are of course many solutions that manage surveys and many other things too.

  • You should embed the rating question in outbound emails to customers. This will give you double to triple the response rate, compared to sending the same survey with just a link in the email. There are various reasons for this. Given the number of phishing and ransomware attacks we hear about, people are reluctant to click on links they do not absolutely trust. It helps if the email comes from your company email server and that this is clearly visible.
  • You need to be able to do a first-pass analysis of the verbatim answers without human intervention. More on this and how to select such software in the next chapter.

While many vendors push it hard, I am not a fan of encouraging customers to respond on a phone. The main value you get from NPS research is the answers to the open questions. I feel you should tolerate phone-based text input, understanding that it will be brief. Your outbound email should say something like “You may be most comfortable providing feedback on a PC, Mac or tablet.”

Next time

The next article will be on how to choose text analytics software.

 

As is often the case, the above is a slightly-edited version of a chapter in one of our books; in this case Net Promoter – Implement the System All of our books are available in paperback and Kindle formats from Amazon stores worldwide, and from your better book retailers.

Operational and analytic infrastructure 800

NPS (28) – The team(s) and tools you need to be successful – 28th article in a series on the Net Promoter System®

/ , , / , , , , , /

Welcome to the 28th article in my series on the Net Promoter Score and System. This time I introduce the part of the Net Promoter System framework that is called the ‘Robust Operational and Analytic Infrastructure’, and go on to provide some suggestions about the team or teams you need to be successful. You can read all of the prior articles on our blog page here.

It is about the team and the tools

This part of the framework is about the team and the tools. What sort of team do you need? Where should it report? What tools do you need to conduct research and understand the results? The Net Promoter System research process produces lots of written answers from customers, and indeed from your own people if they interview customers. What can be used to transform all that unstructured information into deep insights? Are you better off with a set of individual tools, or a single comprehensive solution?

Software is in constant evolution

Software is evolving at a much faster rate than the Net Promoter System. The software that is improving most quickly is that used to analyze open text responses to surveys. Traditional analysis tools require data to be structured. Text responses are unstructured data. Relatively new techniques such as Natural Language Processing are required to deal with it. I will provide guidelines and an example of how to choose text analysis software. The guidelines should not change, but the answer they provide you will change, as each vendor improves their product.

A second area that is in constant evolution is comprehensive software platforms to manage everything from a customer list, to surveying, routing and tracking follow-up actions, reporting and project management. Some software solutions are integrated into Customer Relationship Management systems like Salesforce.com, and some are standalone. I will only cover this area superficially.

Some NPS areas are mature

There are parts of the system that are mature, and where software exists that does a perfect job. This is the case for the basic research or survey software. Within reason, it is hard to go disastrously wrong. I have my personal preferences, which I explain.

Team(s) – Focus produces results

It is easy to say, “Everybody is responsible for the customer”. The problem with this statement is that when everyone spends a little of their time on improving things for customers, but nobody does it full-time, nothing is likely to change. If you are already the loyalty leader in your industry, that may not be a problem in the short term. Success in improving customer experience requires a dedicated person or team. That leader must be well-connected and respected in your organization. Where the leader reports does not matter much. The teams of which leaders are members matter a lot.

Relationship with sales leadership is critical

If you are successful in measuring and improving customer experience, you will sell more. Your sales team should find that value proposition quite compelling. The customer experience leader needs to have an excellent relationship with sales and should be part of their leadership team. I am not suggesting that he or she should report to the sales leader, simply that they must be in the team. My own approach has been to start by proposing to attend an internal sales meeting to present competitive benchmark data. Sales teams find the insights about their competitive advantages and disadvantages to be really useful, so they have always been attentive. Using that positive feedback, I then propose to be a permanent member of the sales leadership team. That has usually worked for me.

Reporting lines in larger organizations

In a large organization, each major business should have a customer experience leader. The job title does not matter. Customer Advocacy is used quite often. Personally, I have held a full-time customer experience role reporting to the EVP of Software at HP and have previously combined the job with the Chief of Staff role for other business areas. The Chief of Staff tends to be well-connected, neutral, and controls the leadership agenda. Having editorial control means customer experience can always be part of strategic discussions. If the Customer Advocacy leader is a full-time role, a great relationship with the person setting leadership meeting agendas is important. Customer Experience falls into the general category of things that are important but not urgent, and can easily become an agenda also-ran.

Chief Customer Officer role

There are aspects of customer-centricity that cannot be addressed within a single business or function. We have seen the rise of the Chief Customer Officer (CCO) over recent years. The CCO should report to the CEO. The job description should include most of the following areas, depending on the size of your company:

  • Chair and run a cross-organizational Customer Experience Council that shares successes and agrees strategy across your company.
  • Define, fund and implement the on-boarding and refresher NPS training that should be mandatory for all employees of all levels, including the executive leadership team.
  • Define and implement appropriate NPS measurement and reporting standards across the company.
  • Regularly report progress to the Executive Leadership Team.

The only advantage I see with the CCO title is that it is better recognized externally than, for example, the Customer Advocacy Officer title. The job content is the same.

Customer complaints team

It is a good idea to have a single customer complaints team for your entire organization. The phone and email contact details for complaints should be visible at the top level of your company website. “Log a complaint” is a reasonably clear heading. The person or people should accept complaints from both internal and external sources. It is difficult for complaints team members to be effective if they are new to the company. The ideal profile is a relatively experienced company veteran who understands the organization and can get things done quickly. A single person could do this on a part-time basis in a small company. A team is needed for a large organization. When I managed this for HP in EMEA, covering consumer and enterprise businesses, there were over 100 people in the complaints team in Europe alone. We had 350,000 employees in the whole company at the time. The range of work was quite extreme, even including representing the company in small claims courts when people went there to get satisfaction on a printer or PC problem.

The complaints team, and perhaps your full organization, should have the authority to spend a certain amount of money to resolve complaints immediately, without further approval. That amount should be something between $1,000 and $10,000. At HP, the Swiss complaints team could leave the office, go to a local electronics store, buy a product and drive it to the customer’s home near Zurich, if that was appropriate. Flowers and other small gifts are a good way of apologizing when you make mistakes or a product is defective. Think about this when you are trying to turn around detractors. Note in passing that most countries have rules prohibiting gifts to government employees, so your processes need to take this into account.

Next time

The next article will be about the survey infrastructure you need for success.

 

As is often the case, the above is a slightly-edited version of a chapter in one of our books; in this case Net Promoter – Implement the System All of our books are available in paperback and Kindle formats from Amazon stores worldwide, and from your better book retailers.

< 1 2 3 4 5 >»

↑

Net Promoter™, Net Promoter System™, Net Promoter Score™, NPS™ and the NPS™-related emoticons are registered trademarks of Bain & Company, Inc., Fred Reichheld and Satmetrix Systems, Inc.
CustomerStrategy.net is owned by Maurice FitzGerald Consulting, Geneva, Switzerland • Site developed in conjunction with iCulture and iBusiness

Show Buttons
Hide Buttons