Quantitative research & evaluation methods:

Data mining

Much of what we do involves asking your audiences questions, for instance in surveys or interviews. We call that “primary” research. But sometimes valuable insights lie waiting in information you’ve already collected. We can analyze your organization’s databases of transactions with subscribers, ticket-buyers, members, students, donors, alumni, and other audiences. Analyzing or “mining” that existing data, with or without overlaying additional third-party data from the US Census or consumer profiling companies, is called “secondary research.” It can reveal latent stories about how your audiences engage with your nonprofit.

Data mining can help identify behavioral segments within your audiences; create a model of how participation grows over an individual’s lifetime; spotlight populations at risk of lapsing; target your communication investments more cost-effectively; and suggest new marketing or retention strategies.

Data mining is sometimes used because it’s less expensive than conducting a new survey or qualitative research project: you already have the data, it just needs to be analyzed. In other situations, data mining is the first step in a multi-phase research initiative, a way of learning “the basics” about your institution’s audiences before lauching into primary research. Ideally, though, the two work hand in hand: we combine individuals’ responses to our survey questions with their actual purchase, enrollment, or donation histories from your database. By analyzing both what they say and what they do, we create an action-oriented picture of not only what’s going on, but why.

Like all research projects we undertake, data mining at Slover Linett begins with a conversation about your goals, hypotheses, and questions. We'll work with your IT wizards to get an electronic file of all pertinent information about the selected patrons. This could include purchase behavior, donations, attendance, first membership date, zip code, and many other “fields.” We then analyze the data using segmentation techniques, regression models, geodemographic mapping, and other tools.

re:search newsletter

More info

Keep in touch. Sign up for our monthly e-newsletter, re:search, and be the first to know about our reports, articles, professional dialogues, and more.

Our blog. Your comments. Jump in.

March 14, 2014 | Nicole Baltazar

Multiculturalism is key for creating inclusive arts experiences


Last month, Coca-Cola aired its now-famous Super Bowl ad depicting people from various racial, ethnic, and cultural groups singing “America the Beautiful” together in different languages. Among the instant outpouring of polarized reactions to this ad rang much praise for its depiction of a multicultural America. Yet the ad provoked a slew of negative responses as well. Many of the ad’s detractors questioned whether this multicultural America could ever feel as cohesive as an America whose citizens speak a common language, and therefore have taken great strides toward assimilating into a common culture.

More »

In practice...

We’ve used data mining to... 

  • discover trends and patterns in a university’s alumni donation histories
  • develop crossover marketing strategies for a consortium of arts organizations
  • gain insights for a symphony about why some patrons don’t return (a “churn study”)
  • combine several years of a graduate school’s admissions data to learn more about who applies and what kinds of outcomes fend to follow

...and many other arts and education institutions.