Avoiding a costly “metrics obsession”

This year, 2011, is the centennial anniversary of the publication of Frederick Winslow Taylor’s autograph work, The Principles of Scientific Management. According to Taylor, almost every challenge management faced could be solved through the application of science. This view has become the staple of business schools for the better part of the last century, as a result.

Most small businesses—which, by the way, constitute the majority of all businesses in the U.S.—found the application of “scientific management” to be unduly burdensome. Many entrepreneurs lacked the training in the application of statistics or the time and energy to conduct “time and motion” studies when they knew—by the proverbial “seat of their pants”—that they could make a profit if they took this action or that one.

By the middle of the 20th century, another great voice in “scientific management,” W. Edwards Deming, was beginning to clear the air on the subject, a bit. While Deming certainly believed in gathering data and analyzing statistics in order to improve operations, he was also unequivocal about the limitations of “metrics” in achieving business success.

It was Deming who pointed out, for example, that “The most important figures for management of any organization are unknown and unknowable.” (Emphasis added.)

---

“The most important figures for management of any organization are unknown and unknowable.” – W. Edwards Deming

---

However, in the 1980s, along came the introduction of the “Personal Computer” (PC) and a plethora of software that enabled small businesses to collect, analyze, store and recall hundreds of thousands or even millions of data points. With the growth of computing power and falling costs of computer hardware and software, the collection of volumes of business data was soon within the reach of even the smallest of small businesses.

Even before the dominance of the Internet as a means for sharing data and collaborating across huge distances, many small-to-mid-sized business executives and managers had become enamored with the ability of computers to store and retrieve data. Even if they were entirely unaware of the pronouncements of Frederick Winslow Taylor, these executive and managers came to believe something along the lines of: “If we can collect and access enough data about our operations, we will be able to manage flawlessly.” The obsession with metrics had, indeed, come of age.

---

The mantra of the “Obsession with Metrics” crowd: “If we can collect and access enough data about our operations, we will be able to manage flawlessly.”

---

Another all too frequently heard proverb from the metrics-obsessed crowd is this: “You can’t manage what you can’t measure.” This, of course, has a tincture of truth to it, but is misconceived. There are all manner of things in which management is involved in “managing” in some way or other that are not not subject to objective quantification.

Here is a (non-exhaustive) list for your consideration:

• Corporate culture
• Customer relationships (we even have software that is supposed to do this!)
• Employee relationships (we have both software—human resource management applications—and entire third-party firms that engage in this kind of “management”)
• Customer loyalty (some companies even have “teams” or “departments” engaged in “managing” this aspect)
• Creativity / innovation
• Leadership
• Ethics
• Supply chains (especially the ‘relationships’ that really make them work; not just the inventory ins-and-outs)

Now, let me very clear here: I do not oppose the application of sound scientific principles to business when the application of such principles is done in an environment where cause-and-effect can be reliably demonstrated.

The correct statement is this one: “If you cannot define the ‘process’ and the theory underlying the cause-and-effect relationships in the ‘process,’ then you cannot manage it.” More importantly, if your theory is wrong, you will not get the results you expect.

---

“If you cannot define the ‘process’ and the theory underlying the cause-and-effect relationships within the ‘process,’ then you cannot manage it.”

---

This clear and correct statement explains why some companies actually see significant improvements in their business results after implementing new supply chain “management” (SCM), customer relationship “management” (CRM) or human resource “management (HRM) applications” while the vast majority of companies see little or no improvement.

Understanding your existing business processes (hint: it is likely they are NOT what you think they are) and tying them to a theory that will help you understand the cause-and-effect within your processes is not as hard as it seems. Nevertheless, most businesses fail to do so simply because they don’t know they need to do so! They think they already understand them—but do not.

That’s why no matter how many “metrics” they throw at the problem and—sadly—no matter how much money they throw at “fixing” things, they typically see little or no improvement in the things that really matter—like making more money!

There is a better way!

Business Intelligence and “Tribal Knowledge” – Part 1

Recently I was working with a client and, at the opening of the meeting, I asked the six members of their management team who were gathered around the table the following simple question: "What is keeping your from making more money tomorrow than you're making today?"

Their responses were telling: (approximate quotations)

1. "We don't understand our customers: Who is buying, why they buy, or how they go about the process of getting a purchase authorized."
   
2. "The amount of time it takes us to respond to a lead or prospect. We might get 20 to 300 leads from a trade show, but it might takes us three weeks to six months to get back to them after the leads have gone through all the hands and processes in our organization."
   
3. "We don't have a good way to turn the data we possess into information that would be valuable for decision-making."
   
4. "We don't have a good way to classify accounts in our customer relationship management (CRM) software so we know how to best approach them regarding our products and services."
   
5. "We don't understand the secondary participants involved in our sales process with a prospect."
   
6. "Our customers lack the funds to buy our products."
   

What is interesting about this is that, if we take number 2 out of the mix (this is clearly a policy constraint) the other five responses all have to do with "business intelligence." These folks needed to understand their customers better in virtually every aspect.

Now, in their defense, this firm has a fairly complex sales cycle with, potentially, a number of different parties involved. Here's a brief description of the participants and their relationship to the sales process:

• School District – Usually, it is the school district that will end up "owning" the product after purchase. Frequently it is at the district level, as well, that the purchase commitment must be authorized.
  
• School(s) – The individual schools and school administrators may have an impact on the purchase decision. The school(s) must be willing to take on the product before the school district will authorize the purchase, even if the teacher may have convinced the district administrator and board that it is the right way to go.
  
• Teacher(s) – Teachers function mostly as influencers and catalysts to the sale. The teachers often are sold on the product and then become an advocate to aid in getting the product approved at the school and district levels.
  
• School District IT Department – Since the products generally involve technology, it is not uncommon for the schools' or the district's IT departments to have de facto veto power over any pending purchase of such technology.
  
• Government Programs – Since most of the schools in the U.S. are publicly funded, the funding for many of the purchases flows directly or indirectly from some government program. Such programs often set requirements and seem to have a never-ending series of "hoops" that must be jumped through before funds are made accessible for specific purchases.
  
• NFP or Other Sponsor – When the school districts' ability to access funds for a desired product purchase falls short, sometimes not-for-profit (NFP) organizations become a supplemental source of funds. Sometimes, it is even the NFP, seeking a place for its funds in community projects, that becomes the initiator of the whole process. Other times, the interested teacher may know that the school or school district have no money for the purchase, so he or she will seek aid from a NFP organization simultaneous with presenting the matter to the school and district decision-makers.
  

As you can see, with all of these participants, and no single path for each approach, it is understandable that this organization is discovering some challenges in "understanding" their customers. Add to this the fact that their business itself was changing. They were diversifying from the product around which the business had originally been built – beginning to sell a broader range of related products into the same marketplace.

Tools at their disposal

Now, this firm does have some tools at their fingertips. They purchase the use of data made available from a data aggregator that provides a database of schools, school districts and related parties. Some demographic data is included.

Now, I am not privy to exactly what demographic data is available to them – our discussions didn't go to that level. However, for the sake of this discussion, let us say that they have just the following data points for each school and district (in additional to standard data like addresses, phone numbers, and so forth):

• ZIP code
  
• Number of students
  
• Number of teachers
  

Using a tool as rudimentary as Microsoft Excel's OLAP capabilities, it would be relatively easy to spot correlations in the data between product sales (by dollar or by units) and these demographic characteristics:

• Which regions of the country account for the most sales? The least sales?
  
  • Using the first digits in the ZIP Code gives you 10 regions automatically
    
  • Using the first three digits in the ZIP Code gives you a breakdown by what the USPS call SCF (Sectional Center Facility)
    
• Which states account for the most sales? The least sales?
  
• Which cities account for the most sales? The least sales?
  
• Which districts produce the highest ratio of unit sales to students? Which ones have the lowest ratio? What about the unit-to-teacher ratio?
  

My guess is, if they had graphs of these data – especially TOP and BOTTOM data – their sales and marketing personnel would immediately begin to see some patterns emerging.

Likely, however, they have other data already in their possession that would give them additional insights as to sales patterns leading to a better understanding of their customers' behavior. Take the following examples:

• Which salespersons produced the highest sales in terms of dollars and units? Which ones produced the least?
  
• Within each salesperson's sales, are there significant differences by sales by geographical region or SCF?
  
• Are there correlations between salespersons' sales and the discounts offered? (This would be an indicator of price sensitivity and should be correlated by other factors, like geography or average sale size.)
  
• Do correlations exist between salespersons' sales results and the products or product configurations they sell most frequently?
  

Little of this kind of analysis was being done in a formal way at this firm. However, it seemed that they already knew they needed to "understand their customer" better. They just had not yet thought about how to leverage what they already had in their hands in order to begin segmenting their market and understanding the factors leading to less success or more success (read: Throughput).

We will talk more about this in the next post in this series.

©2010 Richard D. Cushing