Anyone who has ever been involved in implementing a Business Intelligence (BI) solution has also been involved in understanding the costs of that solution. In my time at Disney and AOL, I built, bought, and implemented a number of BI platforms. Total Cost of Ownership (TCO) and Return on Investment (ROI) calculations are a normal part of that process, both to justify the project initially and to ultimately measure its success...or failure.
As both a buyer and seller of analytic technology, I know from first-hand experience that calculating the TCO and ROI focuses on two areas: direct, up front costs, like licenses and hardware, and ongoing costs, like the number of people required to maintain the solution. Though these are both very important metrics that should be a part of a purchase decision, they don’t tell the whole story.
BI teams also need to consider some of the costs that they may not even be including: What projects are our BI analysts not doing? What questions are decisions makers not asking? And really, how do we do it?
What We’re Not Doing: Opportunity Cost
One of the most straightforward, but often overlooked, hidden costs of BI are the opportunity costs. Simply put, an opportunity cost is the cost of what we’re not able to do because we decided to do something else. For example, if I invest money in something that returns 5% instead of another option that would have returned 7%, my opportunity cost is 2%. Because I chose the first investment, I missed the other opportunity that would have paid more.
When we undertake a BI project, we make a decision to spend more than just money—we’re also dedicating time and people on a specific course of action. It’s important to consider what we’re not doing during that time with those resources.
And though this can be hard to measure in any project, I find it easier to measure similarities between alternatives. For example, if project #1 takes a month to implement while project #2 takes 6 months—what’s the opportunity cost of that extra 5 months? What’s the value of what I’ll be giving up because the people are otherwise occupied during that time?
Although it may be hard to identify the missed opportunity let alone the value of that opportunity, it’s important to attach some kind of number to it. It’s certainly not $0.
What We’re Not Asking: Cost of the Unanswered Question
A successful BI project involves people in a lot of different roles, and even after the project is released many of those roles still play a significant part. But as an end user I’ll only enter into that process if I think the question I want to answer is worth the time and effort.
We assume the new BI solution will inherently be better, why else would we devote time and resources to implementing it. But is it answering the questions your end users have? Consider the value of the question not asked because it wasn’t worth the effort. Or worse, if it is worth it, but the user doesn’t have time to wait, so they make a decision based on instinct instead.
As a BI practitioner I rarely have visibility into these missed questions until I talk to the end users. And I’ve seen how often this crucial step is missed. The cost of an unasked question can be significant, for example if we’re deciding whether to extend or pull a marketing campaign, or whether to offer a discount to an important customer.
How We Do It: Cost of the Workflow
Instead of thinking about BI from only a technology perspective (How much is it and who do I need to run it?), we should also think about it from the perspective of an end-user’s workflow. What’s the process to get an answer to a new question? How do I request a new report? What happens after I do that?
These are the costs that will repeat over and over again, and directly impact the other indirect costs we’ve already discussed. We’ve captured the cost of not asking a question above —but what’s the cost of actually asking a question and getting an answer?
If we’re lucky we won’t make another BI platform decision again for many years, but the workflow to get a new answer will be repeated frequently. Examples of considerations here are the number of people required to fulfill the request, including their fully-burdened cost, the cost of systems needed to manage the requirements and fulfillment process, and the cost of end-user time to perform user-acceptance testing.
So what does it all mean?
As you can see, there are many sources of cost outside of the up front cost of procuring and managing a solution. If we accept that providing data-driven insights is an important function—something that I hope you’ll agree with—then you can understand the significance of appreciating the true cost of your BI solution.
We don’t need to try to measure each of these costs as precisely as possible. It’s much easier and more effective to understand the differences in cost between alternative solutions. Every option will have an opportunity cost, how much larger is that cost between the alternatives? How do the typical new request workflows differ between solutions? And even if you’re not actively evaluating BI platforms, it’s a best practice to understand the hidden costs of the status-quo.