Data Analytics - It's who you know that counts!

We're often asked what's our process for analyzing a customers data and helping them to prospect for new customers. It's one of those things where there is no precise formula. As much as I would love to tell you we have an exact science, the truth is that it is more of a balance of science and art along with diligent testing and retesting of your hypothesis.

So take for example a recent project we did for a summer camp.  The project starts with them setting a goal of what they want to achieve in their recruitment campaigns.  We then look at how much of that is going to come from print to web (or direct mail as they once called it) and what we can do to ensure we hit the goals set.  From there here's the process we followed:

  • Analyze all current customers to fully understand their demography and geography. Look for trends, append prizm codes, segment the data and look for more nuanced trends in the subsets. Then start to ask 100 questions about the data that you have:
Did they respond online or offline, when did they respond, what ages where the children that attended camp, Where they first timers or returning campers, How many others from their zip code attended, where they affiliated with a church or school, how many males versus females and so on and so on.
  • Once we built a clear profile of their current customers then we started the search for new prospects. We use a multiple of data sources so that we can cross reference the data and make sure it meets our criteria.
  • When we know what the universe of prospects is we start to segment them into the various groupings for different messaging related to either their profile or the offer we wanted to provide them. We also ensured there were test groups that represented new offers, demographics and geography we wanted to grow in.
  • Then it comes down to deciding on budget and projected response rates.  That can be the most difficult part when you either need to reduce your list to meet budget constraints or expand your list to get the return on projected response rates.  This is where the art comes into play.  In this instance we surveyed a number of individuals to make sure they matched what the data profile was telling us. A fun thing that we discovered in this process, that might lead to a future key profile characteristic, was families with a second (vacation) home might be less likely to send their children to summer camp. We aren't sure if it's a key indicator yet but there was enough feedback to say it deserves more attention in our next program.
  • And finally when all that is completed we do the most important thing of all, we make sure we have comprehensive tracking and reporting in place for all mail pieces and response mechanisms.  Analyzing our results and feedback in realtime will allow us to react immediately in the middle of a campaign.

And there you have it, that's how it all works. This particular project is currently live so hopefully we can report back on our results in a future posting.