Sometimes the best process is no process at all

Continuous improvement is a common catchphrase. It is easy to believe that making every existing process more efficient should be the focus. But what about eliminating the processes altogether?

HP Lean Six Sigma training

My team was responsible for the formal regional Lean Six Sigma process improvement training at HP. Like many large companies, we designated the various training levels by using judo-like ‘belt’ colors. The annual Black Belt training class lasted two weeks and was always held close to Lyon, in France. Twenty to thirty ‘Green Belt’ process improvement experts took their skills to the following level. My job was to provide a motivational speech at the start of each session, emphasizing the importance of the work, and how it fit the corporation’s overall strategy. The part about ‘no process’ sometimes being the best solution went something like this:

“Let’s talk about a specific example of process improvement. I want to talk about a process you have all followed recently: the travel approval process you all used to get here. First, I would like to ask all those who needed four or more people to approve their travel to raise their hands.” About half the class would raise their hands at this point. “I have a question for all those who raised their hands: does HP policy require you to have four levels of approval to travel here?” The unanimous answer was always in the affirmative. “OK, now, how many people needed three levels of approval to travel?” About a quarter of the class would raise their hands. “So, four-level-approval people, did the three-approval people violate a company rule?” Following a short discussion, I would ask for the people who needed two approvals and one approval. There were always some of each.

The discussion usually got quite intense at this point, and it soon became even more interesting. The training host, Paul Maguire, were in my team, as well as the main trainer, John Holland. “Two people have not raised their hands yet. Paul and John, how many approvals did you need?” “None.” “That’s right people. There is no HP policy that requires approval for travel. Your management chains have invented processes that are not actually formal requirements. I trust my team. Now… let’s have a coffee break.” A chaotic and loud coffee break always followed.

Not just theory

When I joined the corporate software leadership team, I was happy to discover that there was no formal travel approval process for the 13,000-strong organization. Trust ruled, and the senior leaders were expected to live within their overall cost budgets. Individual cost line items were strong suggestions. Mike Salfity ran the Exstream software business within the HP Imaging and Printing Group. IPG had a detailed and difficult travel approval process, one I felt was mainly designed to discourage people from even asking. Exstream was transferred to the software division. The head of software asked me to hand-hold the transfer process. Mike phoned me to find out how our travel approval process worked. Here is how it went:

“Maurice, I have a big spreadsheet of travel requests here. What should I do with it? Send it to you?”

“No Mike. Just delete it. We trust you.”

“… … [Lengthy silence] … This is the happiest day of my professional career.”

OK, that example may be entertaining, but does not represent much of a cost-saving.

Eliminating inspection

I learned about process elimination quite early in my career. During the McKinsey Overhead Value Analysis of the Wrangler organization where I worked near Paris, I started to question the value of inspecting completed orders. There were about 20 order pickers in the warehouse. They picked up printed order forms, took garments from shelves and put them into cartons. An automated conveyor system took them to one of three inspectors who checked that the contents did indeed match the orders. I did some random sampling and found that errors still slipped through. After following some order pickers with a stopwatch and noticing a few mistakes being made, I organized a group discussion. I asked whether the pickers noticed the errors. They said they thought they might have made mistakes, but did not check because they knew everything would be checked for them later.

I decided to run a test. First, I put a proper random sampling process in place after the final inspection, just for one week. Then I spoke to the pickers again, saying I wanted to eliminate the final inspection, and trusted them to check their orders if they felt they might have made a mistake. The results were both clear and surprising after just one week. The random sampling process found about half the number of errors, compared to the prior process following 100% inspection. Trusting people, and ensuring they know you trust them, works.

For the sake of completeness, I will mention that I then wanted to better target the random inspection process. Whenever we received a new product line, I had each size weighed and recorded in the product database. The correct weight for each order was then printed on the picking notes. Every carton had to be weighed for shipping in any case. We moved to inspecting only the cartons where the actual and theoretical weights did not match. Overall, we reduced cost and improved delivery accuracy.

A side-note: customers did not care about delivery accuracy

I learned a lesson about this particular cost reduction when I did some customer research. It turned out that none of the store owners I surveyed cared about delivery accuracy in any case. I suppose we were good enough for it not to be a major issue. Their philosophy went like this, “If your delivery arrives on a Thursday, we will try to sell it over the weekend anyway, even if you sent us the wrong things. If we can’t sell it, we will send it back the following Monday.” I certainly won’t go as far as to say that we should just have put random things in the cartons, but this was an area where we could safely cut costs, without taking any significant risk of losing business.

Find out more…

As is often the case, this article is a slightly-modified version of a chapter in one of our books, in this case Customer-centric Cost Reduction, available in print from the main Amazon stores, and in Kindle format from all stores. Feel free to share your comments, experience and suggestions below, or by contacting me at