In my previous posts, I discussed marketing touch analysis and marketing mix modeling. The underlying point has been that you probably already have the data to do this kind of analysis. If you don’t, it might be because you are not collecting it, although it is already being generated by your interactions with customers. Now I want to get you thinking about how much data you need to get started, and to share some insight from industry experts.

During a recent roundtable discussion with a group of B2B marketing analytics experts, participants discussed how much historical data is needed for effective marketing mix and predictive modeling efforts. The answer is relevant to any marketing operations person interested in expanding into analytics, either through building a team of data analysts and data scientists or using the services of outside companies that provide modeling and analysis.

Some participants said that, in order to generate accurate and complete analyses, the minimum requirement is 24 to 36 months of data about marketing tactics, opportunities moving through the pipeline, and deals won and lost. This sounds reasonable, considering that past performance can be a predictor of future performance. However, here are three reasons why you don’t need that much information to get started:

  • Get started. Building an analytics capability is a gradual process, and you need to start somewhere. The big bang approach doesn’t work, because first you have to develop the skills and learn from your experience.
  • Build a constituency. In order to build an effective analytics service, you need satisfied customers. One of the biggest challenges with analytics involves getting people comfortable enough with technology to feel confident using it to make decisions and drive activity.
  • Three years ago is ancient history. The marketing tactics that worked three years ago might not have the same effect today. Your data will never be perfect, and waiting can cause you to miss the opportunity to have an impact.

    • The bottom line: If you want to build an analytics capability, you will need the data to do it. The good news is that you probably already have data to get started. If you don’t, you can start collecting it now while you sort out your analytics plan.