Reporting or Analytics

A question we get sometimes is “I have pretty good reporting; do I really need analytics?”

Good reporting is good. But it’s not really analytics. Good reporting is almost always predicated on your transaction system – it’s usually about “how much of this occurred?”  How many renewals? How many new subscribers? How many by geography?  What was our profit per unit in the last quarter and year to date? 

Reporting has a number of issues.  First, reports have to be written.  If you want the same data as last month, great, otherwise, there’s a report writing task.  If your team is good, then reports are done in days.  If not, usually weeks.  Sometimes longer.  Depends on the report. 

And detailed reports can run to many pages – 20-40 page reports, or longer, are also more than a little discouraging.  Finding the nuggets in 40 pages can be hard work.

And, as we noted above, most reports are almost exclusively historic.

To get around these issues, many people have started to develop dashboards – to allow interactive drilldown into the historic data.  People who have done this swear by it – it brings the data to life, and, if they’re well-constructed dashboards, they let you find the key bits of information you want quickly and get to real meaning and depth.

But this is still “what happened?”

A more interesting question is often “why did this happen?” and the related question “what can we do about it?”  Not only are these questions are more interesting, the bits of data you want to mine to get them are hidden away in your transaction data. These are "behavioural" questions. They get to the heart of motivations. It is it price? Is it the distributor who is being commissioned to promote something else? Is it a competitive product that is somehow superior?  Is it something you did?Breaking that down is a common analytics challenge. 

Consider the question of “why people are leaving” (or churning, or not renewing or not becoming repetitive clients)?  By asking “what characteristics do departers have?” you can open up a whole new line of inquiry.  You can identify the clusters.  You can sort out which ones you would have liked to have kept, and which maybe you're not so unhappy to have lost.  With this you can set yourself up for action.  But this data, though built out of the transaction data, is not easily gleaned from your transaction data. You usually need other bits, and some serious manipulation. You need to take the transaction data, organize it to answer this question, and then go to work on it. Then you can answer questions like is it age-related? Price-related? Gender? A certain product that is losing its appeal? Is it more pronounced in a certain geography? Is it more challenging in some distributors than others? 

The answers turn out to matter, because if it’s age/product and you play with price, you may put a lot of effort into something, compromise your margins, and have limited impact.  Similarly if it’s in a specific geography, is it more challenging competition there? Or something else?

These kinds of answers can lead you to direct actions that are simply not available from “reporting” solutions.  You can get all the reports you want, but mostly what they’ll tell you is what happened, not why, and certainly not lead you to what you can do about it.

The chart at right shows some of the layers above reporting.  Many people go through "reports" to extracted spreadsheets to dashboards - all valuable.  But mastering the advanced tools allows you to "mine" the data for the valuable nuggets. To start to find out the "whys" and to start to plan the "what we can do about it" all based on data instead of intuition. 

To learn more about our Analytics offering, go here.