Getting to Actionable Insight

“Actionable insight” is one of the currently popular phrases in the IT lexicon. However, actually getting to the point of actioning (ycch! Not a word) insight is not trivial. This was brought home to me during a call with Sparkling Logic. This company provides decision management software. That is, it offers a platform for operationalising insight. And, like the business intelligence and analytic products whose results it leverages, it is a self-service offering intended for use by business analysts. My colleague Simon Holloway will no doubt be blogging about the product in more detail but during the course of the discussion the company suggested that decision management and business intelligence should, in principle, merge. Or, at least, converge.

I agree. But that isn’t the whole story. Before you even get to the point where you can derive insight you have to prepare the data for analysis. This has been a big story over the last 18 months. In my forthcoming update to my existing Market Update on data preparation tools I now have more than 20 products. When we published the Market Update a year ago, I had identified just eight products in this space. To be fair, I missed one or two, but that is still a doubling of products in just twelve months. And there are yet more companies planning to enter this space during the course of this year. Now, this may be a fad but I don’t think so – I think it is responding to a real need. However, if it is true that decision management should be converging with analytics then it is also true that data preparation should be converging with analytics also. Indeed, in some cases – but not usually – it already is.

However, this still isn’t the whole story. It’s all very well being able to do data blending and data preparation. But that pre-supposes that you know what’s in your data lake. Or, for that matter, what is available within your traditional, relational systems (not to mention Excel) and what is available out there in the big, wide world of the Internet. You need to know what data might be useful and where it is before you can start to prepare it. This has given rise to yet another class of self-service products for business analysts: data discovery tools.

Now, data discovery has been around for some time as an IT capability – for example, Global IDs specialises in this sort of thing – but it is only recently that we have seen this emerge as a self-service capability for business analysts. Notable vendors include Alation and Waterline.

So, now we have four distinct classes of products all aimed at business analysts and all providing self-service: data discovery, data preparation, business intelligence and analytics, and decision management. I am sorry, but however worthy these individual product categories are, this is nuts. More to the point, it can’t last: all of these product categories must converge. I cannot believe that business analysts want to go through four separate products to get to a final result.

My actual guess is that, despite it being my starting point, decision management will be the last to be brought into the fold. This is because not all insight is actioned by being operationalised. On the other hand, I think there will be two trends. Data preparation vendors that are essentially data quality and data integration suppliers will extend their capabilities into discovery, which will force pure play data discovery vendors to add data preparation capabilities in order to stay competitive. On the other hand, business intelligence vendors will add data preparation. Actually, this is already happening. Adding data discovery will be next. Who survives? Who knows? The one thing I do know is that this proliferation of tools does not help. When you go to a supermarket for self-service, you want everything in the one shop: you do not go to four different supermarkets.