Helping to bring quality education materials and services to schools for over 30 years.

Matching and Modeling

Tools to Improve Your Marketing Results

Jan 10, 2019

By Bob Stimolo

Based on our quarterly survey, most school marketers would say that the most effective media for growing revenues are sales representatives and direct mail.  Fielding a sales force and creating direct mail campaigns are expensive propositions.  Today, we have available to us a number of tools to improve our chances to have a successful return on investment from our marketing efforts.  For many school marketers these tools are unknown, misunderstood, or more importantly, underutilized.

Our industry is fortunate to have a number of school and district databases available to us.  In fact, one could argue that these databases may be too complicated for the average school marketer.  For example, there are more than one hundred ways to target educators.  You can qualify schools and districts by more than 50 different criteria, and you can qualify educators by another 50 different criteria.  That means you have up to a total of 2,500 possible ways to select educators in schools and districts.

Prospect More Heavily in Customer Institutions

Fortunately, there are tools to improve our targeting.  One simple approach is to “mark” the schools and districts that purchase our products and services.  This enables us to select educators in those institutions either as leads for our sales reps or targets for catalogs and direct mail.  We can also distinguish between the educators who are our customers and other educators in the same institutions.

Often, for most products and services, “other” educators in our customer schools and districts will prove more responsive than those in other institutions.  Why? If the institution has purchased from us, we are an approved vendor; at least one educator is using our product or service and can act as a referral; and lastly, the product or service is present in the institution and available for other educators to sample and review.

We call this process of marking schools and districts that are our customers “matching”.  We can simply match on the basis of whether or not a school or district is a customer or we can match on the basis of how “good” a customer that institution is based on how much they purchased from us or how recently they purchased.  While this matching process is a better way to target schools and districts it does not differentiate schools and districts relative to their return on our marketing investment.

Which Institutions Are More Profitable?

A regression model is the tool that will enable us to determine the profitability of marketing to a particular school or district.  It will also predict the likeliness of our profitability in marketing to schools with whom we currently do not do business.

A regression model seeks to find relationships between our return on investment from our customer institutions and the data that defines those schools.  It determines which data has a statistically significant relationship with the institution’s profitability and to what degree.  Then, it uses that data to rank the schools and districts that are not currently customers and puts them in priority from most likely to become profitable customers to least likely.

Without a regression model, we must choose our own data elements in order to select the prospect schools and districts for our promotion campaigns.  However, we can only treat the importance of each field of data equally whereas a regression model weights the importance of various fields of data.

Data Fields Are Not of Equal Importance

For example, if we decide that the expenditure per pupil and the enrollment of a school or district are important in qualifying prospect schools, we have to treat the importance of each of these fields of data equally.  We cannot designate that expenditure per pupil is twice as important as enrollment when choosing our prospect schools and districts.  A regression model can.

A regression model analyzes the profitability of our customer schools relative to the data selections available and assigns a score for each data variable it considers to be statistically significant.  Then, it adds the scores for each data element and ranks all the schools and districts from best to worst based on their total score.

Marketing Tools Can Make a Difference

Among the reasons that relatively few school marketers use regression models are that they are complicated and based on statistical mathematic theory.  I work with several models and have for a number of years so I have come to accept the concepts that drive them.  While these models cannot actually predict the future, they do improve one’s chances for success.

They are particularly useful in these times.  Regression models tell you which schools and districts you should include in your marketing campaigns and which you should not.  If you make multiple promotion campaigns throughout the year, they enable you to vary your penetration of the market based on the statistical likeliness of financial success.  Finally, these models are remarkably affordable, in most cases adding just pennies to your promotion expense per prospect.

You can participate in SMRI’s Quarterly Survey of School Marketers.  Only participants receive the results of the survey.  Simply go to to sign up.