Self-service BI (or self-service business intelligence) has been a buzzword for a while. If you Google “self service BI,” you get 383,000,000 results.
(Interestingly, if you Google without the dash — “self service BI” — you get 409,000,000 results. I can’t tell you the significance of this. Although Word tells me it needs a dash.)
But you don’t Google search results to tell you that users don’t want to rely on IT to get the “data they need from the software they already have.”
So, as a consultant, why would I want users to do more of the work?
Cynically, you might think I’m being predatory. In many places, self-service BI is the next generation of too many Excel spreadsheets linked by VLOOKUP (for which, see of one of my favorite posts – The Pivot Table Gateway Drug).
But I’m not being predatory. I want my clients to succeed. And while (once upon a time) I used to enjoy being the hero and cleaning up other people’s messes, that time is long gone.
I like self-service BI because, done correctly, it means I don’t have to do things I don’t like to do.
I don’t have to spend a lot of time making things look “just so” or creating fancy visuals, which are things that, quite honestly, I don’t do well.
Indeed, my friend Josh says that if you want to see me sweat, ask me to make a Visio flow chart. I’ll get it done, but I won’t be happy.
The key, though, is understanding what self-service BI actually is. And an excellent way to understand this is to compare two web pages from Microsoft: 1) a marketing pitch for Power BI and 2) an architecture diagram for Enterprise BI.
As I go through this, I want to be clear that I’m really loving Power Pivot and Dax. (I don’t do as much with Power BI itself as not all my clients use it and, as I said, I’m not the visuals guy.)
Let’s start with the pitch on the Microsoft Power BI page. There’s lots of good stuff there, but I want to point out three particular comments that are most relevant to my case:
- Explore data easily by using conversational language and get meaningful answers to data questions asked in your own words.
- Get insights instantly from your favorite applications using pre-built data visualization and report templates.
- Get the answers you need quickly using the skills you have today — whether you are most comfortable using other BI tools, Excel, or Azure.
The promise is great. Things are easy. You can use Excel. You can have prebuilt visualizations, etc., etc., etc.
Now, let’s look at another Microsoft page — this time for Enterprise Business Intelligence.
While this page talks specifically about Azure Synapse, Microsoft understands that, even without that step, getting to good BI isn’t a question of mashing it all into Power BI.
You need to consider each part of the process — the data source, how that data gets prepared for reporting and how users reach that data secure. Then, and only then, can you start playing around in Power BI.
Maybe your experience is different, but I don’t think many end users are going to do those first five or six steps.
My point is that self-service BI, when it’s honest and valuable, means that users can work with data that’s been properly prepared by folks who know what they’re doing with data.
And that’s where I fit in. I enjoy getting the data straightened out. I enjoy connecting it all together. I enjoy designing the models and measures.
And I really enjoy watching other folks make it pretty.
So, that’s my idea of self-service BI. Granted, plenty of folks think that accountants who know Excel can do everything themselves. And if they can, more power to them. But most of the time, it winds up being another case of pivot tables gone mad.
If you understand what self-service BI can actually do, then you know that those accountants aren’t going to do the work end to end, most likely.
And that’s where I can help.