Practical Analytics: The Value of Actions
In the analytics world, there are teams that are “stuck.” This article is to help those teams get “un-stuck.” Perhaps you headed down the path of analytics without a plan. Perhaps you had a plan but the ever-changing data privacy regulations halted that plan. It’s time to refocus so you can begin to move forward once again.
Perhaps you took on a new leadership role in analytics, aced the interview but now you wonder what you’ve gotten yourself into. Can you deliver the value from analytics that you promised to your leadership?
Many analytics teams lack business acumen. We’ve said that many times, but what does it really mean? Having good business acumen means having the experience and wisdom to see the bigger picture and to be able to EVALUATE (in the case of analytics) whether a project is worth doing. Too often, we are still seeing THEORETICAL analytics rather than PRACTICAL analytics.
The Value of Actions
Practical Analytics
Assessing whether a project has value should be one of the first filters applied to project ideas. It doesn’t have to be the full ordeal of putting together a formal business case, but there should be some upfront thought about value.
Here are a few questions you can ask yourself…
Will this project improve our competitiveness?
Will this project save us money?
Will this project generate incremental profit?
Will this project improve the speed of crucial processes?
For example, if an analytics project has the goal of improving employee retention, we know that employee turnover has a cost associated with it and therefore it could have cost-saving value. If we have a high turnover rate, this project may have enough cost savings potential to find enough value to proceed with the project. If turnover is very low, then this may not be a project of value to your organization.
If we have an analytics project to assess the rate at which we sell certain inventory to our customers, we know that being out-of-stock can result in lost sales. This type of project could result in increased profit. How often are we out-of-stock? What are our estimated lost sales? Is that loss enough that it’s worth moving ahead with this project?
Additional Practical Considerations
PRACTICAL considerations go beyond the sample questions above. With so much data available to us today, we also have to consider the ethical application of analytics, especially when the project will impact people.
Does our analytics project adhere to data privacy laws?
Can we ensure the security of data?
Can we act on the results of the data analysis?
Are we willing to act on the results of the data analysis? (This is not the same thing as the previous bullet.)
For example, we may determine that employees that have been with us for 3 years without a promotion are more likely to leave the company. Can we act on this information? Technically, yes, we can take action on this result, but are we willing to do so? Should someone be promoted because a data analysis put a higher statistical probability on them leaving than the coworker next to them? If we act, will we end up in court debating the statistical validity of the approach we used for decision-making? (That does happen for topics like adverse impact.)
Acting on Mankind is Different than Acting on an Object
It is easier to act on the analytical results of physical objects than people. If I determine the variation of ordering of my product sales, I can plan how many of each item to stock on the shelves each month. This is also a statistical probability analysis in that it determines the likelihood that I will have an item in stock when a customer wants to purchase it. However, when the analysis points to decisions impacting people’s lives (promotions, hiring, firing), the considerations required before acting are far more complex.
In short…
Too many leadership surveys are showing that companies are not yielding sufficient value from analytical investments. Let’s change that view with the following advice.
If you can’t or won’t act on it, don’t waste time and money analyzing it.
