Here’s a long-play OSS analytics strategy for you to try

Analytics is a term that has caught fire in IT relatively recently. In many ways, our OSS have been doing “analytics” for years, albeit not necessarily with the same tools at their disposal. If we simplify the term down to the use case of “being able to ask questions of a massive data store” then that’s what OSS have been doing almost since their inception decades ago.

However, most big telcos are carving out (have carved out) specialist analytics teams that perform this role, using data from OSS, BSS and any number of other data sources. These teams are expert at being the manual shim between the questions asked by the business and the data store (as shown below):

One of the things I love about OSS data stores is that every new person who comes onto a project asks different questions of the data and in turn, spawns different insights from it. I’ve just come from a project where dozens of people were asking their own different questions and almost all generated “light-bulb” insights.

But the diagram above alerts us to an important trend that is going to gain momentum in OSS. The analytics team, the manual shim between questions and the database, is going to be increasingly replaced with automated tools. Not only that, but the automated tools are going to be able to learn from questions coming from the business and start forming their own questions, giving predictive and precognitive search value.

This capability is nascent, but the trend is unmistakable. The long-play strategy I propose for you is to start automating the collection of questions being sent to the analytics team and bind it to the response data being returned.

I’m proposing that this capability will not only be a great resource for people to see what other questions have been asked (and spawn their own questions), but it will also act as seed data for the machine-learning engines that the automated shim layer will be based on.

BTW. Let me know if you’re interested in helping to build this “question and answer” tool.

However, I should call out that this isn’t just a “question and answer tool.” I believe that this is the future of consulting. Having been a consultant for quite a while now, I’ve noticed that the best consultants are the ones who are most inquisitive and ask the best questions, either of stakeholders or the data at their disposal. We’re moving into an age of increased data-driven decision making. The “gut-feel” consultants may still have a place, but the “data-driven” consultants will have the opportunity to gain greater credibility. More to the point, it’s the consultants who can form a hypothesis, then ask the right question of the right data that will stand out.

This should give the big consulting firms a significant advantage. The small consultancy firms may have the domain experience just won’t have the wealth of data analysis horsepower available. To combat this power imbalance, perhaps there’s an opportunity to develop an “as a Service” model, where crowd-sourced questions and answers are generated by a tool like the one articulated above….

Perhaps the machine can even start suggesting the right questions to ask in certain situations? Or even normalising the questions so the answers become more consistent and then form the basis of a set of baseline KPIs to compare other client data with… Which then increases the chance of providing data-driven advice, not just gut-feel guesses.