Self-serve BI is unquestionably on the rise—24% of IT departments have deployed it to some degree and 84% plan to invest or deepen investment. (Logi Analytics). But there is a non-trivial data management effort that must precede self-service access to BI. This fact is apparently not very well understood, because we hear complaints about traditional BI tools from marketing leaders all the time. Their stories typically go something like this:
“There’s a big push in the organization for self-service access to our data. The IT department sets us up with canned marketing dashboards through Tableau (or other data visualization tool) and says, ‘Here you go!’ But it’s so confusing that none of us marketers know how to even use it. Plus, there are all kinds of inconsistencies in the data—nuances that none of us can understand. I want to look at total views, but in the tool, I see a metric called View and another called Views and another called Views-web (and, and, and…) Each has different values. Do I want one? Another? Am I supposed to sum them? Getting more data-driven and operating in real time with self-serve access to my data was a nice idea, but there’s no way we can use this.”
Nevertheless, the self-service BI drum just keeps beating. I just googled “self-service BI” and got 12,000,000 web pages and 120,000 news articles. It’s a frothy market—most of that content is commercial content published by BI vendors (e.g. web pages and press releases). Even most news articles are clearly driven by vendor news.
Finally, I did run across an article that discusses the Gartner point of view on the subject of self-service BI. It’s much more in line with what we hear loud and clear in the marketplace, but it was a lone voice. Amidst the ocean of self-service BI evangelism, The Trouble with Tech Integration for Advisers, Vendors quotes Gartner as saying, “Through 2016, less than 10 percent of self-service business intelligence initiatives will be governed sufficiently to prevent inconsistencies that adversely affect the business.”
Wait, wait wait. If I may paraphrase: More than 90% of companies out there have not done the data clean-up required to use BI correctly and they’re likely to be making bad decisions that make the company WORSE OFF because of it.
This quote in the article from Edmond Walters, founder and CEO of eMoney Advisor, really hit the nail on the head: “We didn’t even understand what the demand [for data integration] was until it happened.” This sentiment is right in line with what we’re increasingly hearing from the marketplace. Many marketing leaders are lured by the pretty pictures—dashboards, scorecards and data visualizations. They want to be more data-driven and they want the reporting and dashboard tools that help them to do it. But what they don’t realize (until after the 4-6 month development process they went through with IT just to build out their requirements) is that marketing data is uniquely messy, posing a significant challenge—one that most IT departments are not prepared for.
Of course self-service access to marketing data is not always a bad idea. It makes tremendous sense AFTER good data governance, naming, definitions and clean-up have been instituted. That’s what Beckon does for each of our clients—clean up the data mess and maintain that tight data ship over time—so that pretty pictures actually mean something. Here’s how to decide if self-service access to business intelligence is right for your organization:
Self-service BI is a bad idea for marketers unless our data is integrated properly before it ever gets into the visualization tool. And as Mr. Walters (and the scores of other marketing leaders we’ve met with) can attest, without properly integrated data, our marketing “intelligence” is anything but.