Follow the I4 Process Blog

 Subscribe in a reader

Enter your email address:

“Shelley’s ideas about how to engage leadership and how to get the project to succeed at Dey were important to the success of the team.”

Batch Records Team Member
Dey Pharmaceutical

 

Don’t Accept Garbage In

Dear i4Process,

We are modeling and analyzing our Customer Product Recommendations and Pricing Process.   The better job we do there the more loyal customers we have, and of course that means more profits.

Currently we do some customer segmentation to enable us to see customers in groupings.  We have two major problems with segmentation now:  (1) the segmentation algorithms are complex and hard for many to understand and (2) initial quantitative data consisting of customers purchasing sales, employee size, size of order, start date and customer outputs are not complete and accurate, so we can’t make the most appropriate recommendations.  What do you suggest we do?

Yours,  Open

Dear Open,

There are two techniques that your questions imply so you are already thinking in the right direction.  First, standardization:  if you can create standardized algorithms for the key customer segmentations you will have a simpler way to approach the customer groups.  Hopefully, you can develop a limited set of customer groups (say 5-15 vs. a large set of 100 or more).

Second, get 100% complete and accurate information in at the front end (one of the rules for redesign that i4Process recommends). I use a technique called an Input Evaluation Template to help identify the input requirements.  Here’s what it looks like and how it works.

 

Input Evaluation Template

Let me explain the table.  Start by listing key inputs and put them in column A.  Often the most important inputs are inputs that come into the process right before the process starts, or early on in the process.  But there can be key inputs at later steps in the process.  Most of the inputs for a service or information process will be data/information in some format—on input forms (paper or electronic) or data sets from the vendor or customer, which was the situation in this case.

Now take one input, e.g., customer employee size and work across a row filling in each cell of the template.  Answer the heading: 

What challenges are there with the input today?   (It is often incomplete and inaccurate, or it might be not timely, or not standard, or something else.)

Who is the Provider?  Does the customer, a vendor, an outsourced group, or another department in your company provide the input?

What are Our Requirements?  In other words, how do we need to input to be?  (In a certain format, 100% complete and accurate, no duplications, etc. )

How does the data come in today?  Here you are looking for quantitative data about how  “good” the input is now. Get real numbers or categories with numbers.  This example still needs some work on that.

Ideas from Others.  This column is your opportunity to put in ideas from other companies where you have seen it done differently and better, and these ideas could also be your own improvement ideas. 

Once you have completed the input in the first row, go to the input in the second row and complete the cells working across for that input.  (You don’t need to complete this template for every input, but for the important ones.  You could also sort the inputs by provider.)

Next you have to step back and see what you learned from the template.  What’s obvious here is that getting 100% complete and accurate information at the beginning is critical (even though this company said they would accept 1% errors.)  That would impact the first and second inputs.  The third input problem (time) really needs to be solved within the company with faster servers.

An IMPORTANT WORD OF CAUTION.  Do not go tell your Provider they messed up.  Often the customer may not know what you require or how to give it to you in the best format.  This is an opportunity for a dialog with the Provider – what you need to help them, and what they need to help you.

Want to learn more about analytical techniques to improve your processes?  Register for Analyzing and Optimizing BPM Processes July 30 and 31.  There’s an early bird coupon available now to save 15% off list price.

Be Sociable, Share!

Leave a Reply

  

  

  

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>