Why your data doesn’t work for you
Every unhappy database is unhappy in its own way, but there are some common things that come up.
Integrations are hard.
Often when people say their data doesn’t work, they mean they can’t use their data in situations where it is necessary – for instance, if you can’t email just members who have given over $500 or something like that. This problem is frequently about an integration, and the root of the problem is often that someone assumed that dumping data from one system to another would integrate them. An integration is when you take data from one system and explain it to another system, and so they’re harder than anyone wants them to be.
People also say their data doesn’t work for them when they can’t get accurate, consistent answers from their data. The top 4 reasons I see for this are…
The question is poorly-defined.
People in lots of industries talk about retention rate. The dirty secret is that most people who talk about retention rates can’t tell you clearly what they mean by retention, and they may not even mean the same thing each time they talk about it. There’s more than one way to honestly talk about retention, but if you haven’t defined exactly what you mean, you’re going to get inconsistent answers that feel completely wrong. Comparing retention rate between organizations is even dicier. If you’ve been asked to submit your retention rate for a survey and then had it compared to others, the chances that you’re comparing apples to apples is pretty low.
People hate the answer.
I don’t mean to be flip about this – people sometimes ask really good questions and then the answer winds up being so far from what might be useful that they just can’t process it. For instance, if you want to know how long it takes for someone who reads your content to support you financially, you’re probably thinking that the answer is something like six months or a year or even two years. You could do something with that information! You could guide their journey and try to speed it up a bit, or make the messaging particularly attuned to that timeframe. But what if the answer is 10 years? A few months after you ask this question, you’ll probably have stored the experience in your brain as “the data doesn’t work for me,” and you’re not wrong. The data didn’t help you out in this case, because some information isn’t actionable. It just means it is time to ask different questions!
There’s too much distance between the people who analyze the data and the people who are up to their elbows in the data every day.
In most organizations, there’s a lot of distance between the people who analyze data and the people who really understand the data. They’re often far apart on the org chart, far apart in terms of status and pay, and far apart in terms of life experience and worldview. If nobody forces them to talk to each other, they won’t, and your analysis will turn up answers that are just plain wrong. You might have seen the news in 2021 from a county that reported that 200% of people over 65 had received their first covid vaccine dose. We all know that’s wrong! If you don’t track whether someone lives in the county where they’re getting a shot and which dose you’re giving, you can’t accurately report on this number. Lots of organizations do exactly this thing – and sometimes the answer is just as wrong, but it isn’t as obvious.
Artisanal Data 