Author: b a

1 min read

Is doing something hard inherently more valuable than doing something easy… even if the hard thing is less useful?

A client has an in-house data analytics team. They do really hard things using tools like support vector machines and natural language processing.

And, we heard some backhanded comments about how our approach to solving business problems is a little too continuous improvement focused and not quite cool enough to really be data science. I definitely take that personally: hence this post.

Here’s my observation (full of bias, I’m sure): the in-house team is aligned to a support function; therefore, they take on data science projects that sound complicated and are aligned to the needs of their functional executive. In contrast, we define our project customers as those parts of the business that generate the revenue stream. We do this so we can better understand value. Drawing from Lean thinking within the continuous improvement space, value is defined by the Customer.

No one else.

While we are on the subject, let us review how businesses add value. A business may do lots of activity. But, a business activity adds value if and only if:

• the Customer is willing and able to pay for the activity
• the activity changes the fit or function of the product or service
• the activity is performed correctly the first time

What does this mean? It means we work on that which is important to the Customer. It means we do not subordinate our efforts in favor of displaying our analytic prowess or employing an exciting tool. Tools are simply what we use to get results. They are not the reason for our existence.

Hence our focus on solving business problems tied directly to the revenue stream and our desire to use the right tool for the job.

We had a (virtual) conversation with a prospective customer the other day. The call was largely attended by the customer’s engineering community. We pulled together some preliminary analysis on a development process. It had a very small sample size, which we noted and detailed using appropriately wide confidence intervals. During the call, several of the team members chose to pick on this fact rather than the key conclusion that their process simply was not capable of meeting their customers’ needs. Fortunately, Zeno gave us this playbook long ago:

“We have two ears and one mouth, to listen twice as much as we speak”.

So we kept our cool during the discussion and listened closely as the team indicated that they really needed help looking in a different part of the process.

Bingo!

We politely took an action to address their concerns and dig into the data from this other part of the process. Customer relationship: re-invigorated.

This quick story reminded me of a presentation Dan and I made at the Association of Manufacturing Excellence (AME), one of the biggest Lean conferences in the US back in 2014. I think it nicely illustrates how no one continuous improvement methodology is universal. And, it demonstrates how right Zeno was even about something as arcane as process improvement! Here is the approach we outlined back then. Slightly modified, it still holds for our practice today:

• Listen closely to the customer. Diligently try to understand what problem they need solved
• Review the full portfolio of potential problem-solving approaches
• Select what looks like the best fit for their unique situation
• Get feedback quickly and be prepared to adapt

Sign up below to get the full presentation we made at the AME conference. Enjoy!