A friend of mine is an executive at a smaller non-profit healthcare provider. We’ve been speaking informally about the various data and reporting challenges his agency is facing.
Here are the key facts:
- The organization relies on multiple software systems, some in the cloud and others server-based.
- The organization has several folks who develop reports in the different systems. Each person develops reports as they see fit, generally upon request from others in the organization.
- There is no CIO or CDO. Various other executives (including my friend, who has no technical background) have responsibility for managing technical folks.
- An executive with extensive experience in larger, more sophisticated organizations wants the organization to become more sophisticated. He would like them to develop a data warehouse to centralize all their data so they can better develop metrics that would drive the business forward.
- The organization has not been successful in the past when they have attempted to execute large projects.
What do I recommend? Should they be strategic, develop a data warehouse, create metrics and distribute? Or should they go tactical and identify all current data demands and knock them out one by one?
Technical versus adaptive challenges
If you’ve followed the relaunch of this blog, you’ll see that I’m thinking about and working through the ideas around technical versus adaptive challenges.
My friend first reached out to me because he knows I’ve done a lot of work with business intelligence. Since his organization is talking about reports, he wanted my technical advice.
But the more we talked, the more it became clear his situation is an adaptive challenge. Yes, better data is a “good thing.” However, I’ve built too many reports that didn’t get used to believe that if you build it they will come.
Let’s take a step back. What’s the point of a data project? A dashboard? A well-designed report? It’s to allow people to act differently because of what the data shows them.
Maybe the data shows an opportunity to take, a challenge to overcome, or a problem to solve. But the point is, you should only be investing in data if you’re going to do something with it.
Based on our fairly short conversation, it seems that the folks in my friend’s organization pretty much do what they do every day. And it pretty much works.
It could be better. But when they try to make major changes, they revert back to doing what they did before. And have it pretty much work. Could things be better? Probably. Does the culture embrace change? Not really.
Making a little bet
Going full on with a data warehouse is an expensive consideration. If your folks aren’t currently clamoring for better data, a data warehouse project probably isn’t the best place to start.
A number of years ago, I read the book Little Bets: How Breakthrough Ideas Emerge from Small Discoveries. The point being, you need to give things a try. You don’t know what’s going to happen or what’s going to be useful. Do something small and see how it works. And then let that actual experience guide your next efforts.
In my friends’ case, I’d advise them to choose a particular challenge they think data might fix. They should identify the challenge and what better data will enable them to do. They should figure out how to provide the data in the simplest way possible. (If that means sending out excel sheets for a few months, that’s OK).
They should also be explicit that they’re trying to change behavior. They have an adaptive challenge they need to face.
And then, they should see what happens – which doesn’t mean just send out the reports and wait for people to ask. They should have a plan for working with their folks and their reports and discuss what changes they make.
After two to three months, they should have enough data about the experience to see whether they should go forward.
They’ll have data about how they work with data.
It’s easy to talk about how we think we should work with data, and it’s another to see how we actually work with data.
What’s your experience? When people get the data they ask for, does it actually change their behavior?