Written by Sam Tyler (senior consultant in our healthcare sales & marketing intelligence practice):

 

No matter what organization I’ve worked in, the clamor for data is palpable. Data is all over and there’s a whole discussion about getting huge amounts of data from sources that are tricky. My experience says there’s still a gap in getting existing data in the first place. So how do you make that data accessible?

The preamble – Start simple. That’s not necessarily small. Some of the most fruitful but simple data to gather is claims data combined with member data. Sales channel data with a focus on who gets credit for a sale along with the ongoing customer acquisition work also has many uses. Remember, though, to have the physical architecture. Get some servers; get some databases up and running.

STEP 1: DATA Analysis/Modeling/Exploring. Create a model that conforms to best practices. This should be based on your business, not on your existing system. I continually see data warehouses that conform to what the business has, not what it should have.

STEP 2: INTAKE the data. Just copy it from one source to another. There are all sorts of ways to automate this and reduce the cost and time of onboarding data. Otherwise, you could hold up the production system as well your own.

STEP 3: STAGE the data. Aggregate, reshape, normalize (not data normalizing, that should be part of STEP 3; this is statistical normalizing) – whatever you need to do to the data in order to get it ready for analysis.

STEP 4: COLOR in the blanks. Go back to STEP 3 and find out what’s missing – either new parts of the model or unpopulated parts. STEP 3 becomes your blueprint.

The one thing to remember here is to iterate. Don’t try to get everything at once. It’s better to have one metric used a thousand times than a thousand metrics used zero times. You can have member age, claim count, and plan in a week from the time you get access (assuming you have the preamble done). That’s far better than investing three years into a project that never produces one gram of insight.