Sep 10, 2018
Breaking Down Analytics
The Value Of A Step-wise Approach To Analytic Success
Analytics can be overwhelming. Too often it is the realm of big questions, big data, big projects, and big budgets. All of this requires a big plan and a big leap. It is a big risk. All of which leads most often to big failure.
It doesn’t need to be this way. Unfortunately, it has been this way for so long that the industry almost demands it. Tools, technology, consultants, and other support are all designed for big contracts, big budgets, and big hype. Personally, I think the industry is overcompensating.
Consequently, it takes a lot of confidence, built on a lot of experience to do it differently. That said, it is precisely the process that results in success a lot of the time. Sadly, right now, it would require a lot of research and a lot of networking to find those willing to break down analytics, to transform it from big leaps — to a lot of smaller ones.
Taking a big leap can be fun, as long as it doesn’t end in a face plant. But then there are also broken bikes, broken bones, embarrassing video, and a host of other less-than-preferable outcomes. Is that really any way to run your business?
Let’s step through the alternative.
Much of analytics is analysis (though not all). Yet very little analysis seems to be applied to analytic projects (at least openly). It as if practitioners feel that “breaking it down” or showing the steps in the process will make it less impressive, less magical, or less valuable. If analytics were only analysis, only science, or only that easy — this might be true. But analytics also requires synthesis, it is an art, and it is a field where experience is never optional.
Experience also tells you that you can take analysis too far. It is possible to break things down to a point where little value is added, while greater complexity is assured. Experience tells me that session level analytics is the place to stop.
A session is a targeted activity with a clear purpose, value, and output.
Breaking analytic projects into individual sessions has a pretty stunning impact — it makes big projects less risky, more agile, and (when done right) less expensive. Session are highly iterative and typically end with the delivery of a plan. When I talk with clients about them, they are either labeled by the activity we focus on OR the plan we will developing.
So we might engage in:
- Scoping sessions
- Strategy sessions
- Discovery sessions
- Diagnostic sessions
- Consensus Building sesssions
- Storyboarding sessions
Or we have analytic sessions to create:
- Measurement plans
- Testing plans
- Resource & staffing plans
- Execution plans
- Data monetization plans
- Integration plans
Each broader type has dozens of varieties. I list only a half-dozen or so, but hopefully this list provides some examples of level and purpose of a session. Using a discreet segmentation like this allows for clear articulation of outcomes, focused documentation, and clearer management goals. It can also be very difficult to organize for those with less experience. Analysts that are still in that “hacking phase” of analytics are unlikely to be able to orchestrate something this “clean”. For that matter, analysts strained for resources and drowning in prioritization lists will struggle,too.
It may require a lot of sessions to overhaul a wayward company. It may require a lot of steps to deliver on big visions and dreams. But small steps are easier to resource, easier to plan, and easier on the budget.
Unfortunately, too many in the field of analytics are infatuated with the big trick. It is the only way to ask for big money and they need big money to offset their big risks. Unfortunately for you business, they aren’t the only ones risking a skinned knee.
So, if you are investing your money in analytic support, give it to someone who will invest it in smaller steps. Invest with someone who has a lot of experience and capable of getting a lot of value. Someone secure enough to not need to compensate or show off. Best of luck and thanks for reading.
For assistance:
