One of the hardest things to do in business (specifically marketing) is to data-mine missed opportunities. It’s hard enough to get people to care about your product/service, but it’s even harder to get the ones who don’t really care/want it to tell you why they don’t. There are tons of reasons why you may have missed out on connecting with a potential customer or client, so where do you even start to narrow it down? Surveys? Focus groups? Research firms? Those are certainly possibilities. Usually though they’re either ineffective, skewed due to sample size, or expensive. 

My recommendation? Start by gathering the positive signals within your customer base. Find the people that are hot on your product or service and get to know them through a combination of observation/behavioral data collection and surveys. When people love your product or your company, they are more likely than others to respond to your surveys. Next, follow up on cooling/cold leads (this is a lot easier for software products and service providers to see coming) to see if there are ways to spur them back into use. Use A/B tests to verify that your efforts are really whats driving behavior and not other factors. In other words: Quantify, quantify, quantify. Qualitative data is hard to observe in A/B tests, so focus on what you can observe concretely and avoid inferences where possible.

Next, look into the market to see where your non-customers are investing their time and money. You can often use publicly accessible studies or published data from other companies to understand differences between your customers and theirs (keep in mind that some of their customers might be yours too, so take this into account when reading your data). Investigate commonalities across your customer base and commonalities that exist across the people outside of it.

Finally, you’re ready to use some inferences to fill in the gaps between your data set and the unknown, but make sure that you proceed in that manner- that this is an inference. Too often, product managers, marketers, or other decision makers begin making inferences in step one and as they proceed down the path of research they can forget which information is definite and which is an assumption/inference. When we start to make concrete claims that are truly inferences, the bar for judgement and decision-making gets lower and lower. To avoid this, make your bar for inferences high—maybe higher even than your bar for the truly concrete data. This way you can build a foundation of likelihoods rather than theories. Internal data + market data + public data + intuition & critical thinking = A great foundation for likelihoods. Just don’t skip any of the ingredients.