it all feeds from operations
What is the point of data? What about a report? What about a dashboard?
Let me rephrase my thought here…
If a tree falls in the woods and there is no one there to hear it…
If a report is built and no one uses it to inform a decision is it useful?
I would argue no.
This line of thinking has multiple implications for analytics or BI groups:
1. How we internally measure success
2. How our efforts are delivered
3. Future group organization structures
Are we worth our weight?
I find it interesting that the groups focused on delivering measures and analysis to help other areas of business track and improve performance often do not measure their own efforts effectively or at all. there are plenty of articles that propose a number of attempts at measuring a “return on analytics”, do a google search and hundreds of pages will appear. I have probably read at least half of these articles but none seem to cement the theory for me. I had to step back and think outside the realm of data and analytics to create my opinion.
we measure success by the decisions positively influenced.
You may be wondering how to track these occurrences? An elaborate form based web application directing users to explain how data changed a process or thought? What about a survey that we could run basic sentiment analytics on?
No, how about a simple spreadsheet? I know, so inefficient right? but effective, I think so. But this approach lends itself to the next two pieces of the puzzle.
Ding, pizza is here!
Traditional report delivery.
1. Please find your report attached…
2. Please click the link to view your report
3. Please check your folder to view your report
As the groups building these reports move more and more into the functional areas we should be tailoring our delivery methods to those areas. To track how our data/information is being utilized to drive value we must be in the functional meetings where the processes are being discussed, where the decisions are being made and most importantly (in many cases) where the planing is taking place. We need to be in those early stage environments to ensure that the user cannot escape the data and it becomes a hardened part of their daily process. It is much harder to miss the tree falling if it is already in your way. That is where the information needs to be delivered.
An oversimplified use case…
As a small business owner opens their invoicing software to build another invoice, the customer’s payment / order history is front and center with call-outs to late payment tendencies with small quantity invoices which in turn leads the owner to wait to invoice for a larger quantity saving him the later hassle of collecting late fees.
Does this change who we are or will be?
The standing (paraphrased) definition of a data scientist is someone who is a better at statistics than a programmer, better at programming than a statistician and with greater business acumen than both. So what am I saying, we also need to be experts of process design? And need to be great at sales and influencing end users of our value? And responsible to our leaders for our financial impact? And must run marathons, do crossfit and bodybuilding… wait, too far. but everything before that, yes.
What mighty superhuman being could do such things? None that i have met, at least not well or in any sustained way. Here in lies the value of great product teams.
In addition to all of these functions not living in a single human, they (in most of my experience) do not live in a single business function or under a single leader. this is where the idea of product teams consisting of many stakeholders with a single (incentivized) goal can fill that void.
I will save my product team rant for another day
To wrap up, our jobs as analysts, data folks, report monkeys, analytic squirrels or even the famed data scientists are changing… for the better. As long as we focus with the end results and impact held very close to our goals on every deliverable we will be successful.