Q&A with Jill Dyché: Pooh-Pooh Indeed!
In which Jill considers the end result.
One of the happy discoveries in writing a Q&A column for Upside.com is how impressively smart my readers are! Not quite Neil deGrasse Tyson or Stephen Hawking smart; more Lena Dunham or Stephen Colbert smart. Smart with a world-weary, "We get it, don't we?" wink.
Smart like J., who submitted this question:
I'm not a visionary, so I often pooh-pooh the beginning of important trends -- MTV, cameras on phones, and big data, to name a few. Now I find myself pooh-poohing a certain large vendor's new suite of products. One speaker at this vendor's recent summit claimed that becoming an expert in these tools now is an opportunity to get in on the ground floor of a technology revolution. That kind of hyperbole immediately triggers my pooh-pooh response.
I am writing because my company is reviewing new enterprise BI tools as part of an overall BI program, of which I am the program manager. Because we're implementing a self-service reporting platform, we have a cross-functional business team evaluating tools for ease of use and functionality. There are a couple of evangelists on this team who keep asking why we would bother with another tool when everyone already knows this vendor's historical suite of tools and it can basically do anything.
My argument is that the vendor's tools are file-based desktop tools, not enterprise tools. There's a host of critical governance and administration functions that other BI products handle that this tool suite does not. Absent those management capabilities, we'll end up with spreadmarts with prettier graphs. In fact, the newer, sexier tools didn't have those enterprise features when they started. Now they do, many releases later, after having been beaten into submission by the enterprise BI market.
I like my argument. I think my argument wins. However, I could be wrong. Am I missing the boat again or am I right to pooh-pooh?
--J. in Pooh-land
See what I mean? We need more people like J.
It's a problem at most companies. When faced with weighing in on a new product or tool, many people shrug their shoulders at fit, opting for whatever they perceive will best burnish their resumes.
It seems as if J.'s colleagues are advocating on behalf of the software they're most interested in learning, irrespective of how well it meets current requirements or integrates into the incumbent infrastructure. At best, this is the path of least resistance. At worst, it's a conflict of interest.
However, J. also mentions that the company's BI environment supports self-service. In this case, the answer might not be selecting one product over another -- enterprise-ready or not -- but letting multiple solutions coexist.
After all, as BI problems are different, so are BI solutions. The practice I've seen work well is the "BI portfolio" -- a series of BI and analytics applications deployed across lines of business incrementally and over time.
For instance, your finance department might use Tableau for reporting and data visualization. Your operations group might have embraced Microsoft's Power (fill in the blank) suite, while your marketing team might use the Customer 360 suite of tools from SAS.
Such a best-of-breed approach ensures that so-called corporate standards don't usurp fit, and lines of business -- aka "the users" -- have a stake in technology selection. It also keeps vendors on their toes, knowing that a competitor is only an elevator ride away.
At an automotive company, siloed lines of business had historically acquired their own BI and analytics solutions. Finally a frustrated IT director insisted on holding a "BI bakeoff" to determine a single, enterprise BI solution, which would become the company's established BI standard. The director defended his decision to me, citing efficiencies and lower support costs.
"Is that the hill you want to die on?" I asked him when he finished. He looked at me as if I were wearing a palm frond.
I explained that mandating a single BI solution meant taking tools away from people trying to do their day jobs. Show me someone who wants to dictate a solution to a group she doesn't belong to and I'll show you software that becomes shelf-ware. Heck, I'll show you shadow IT!
So, J., you get it. No one tool will solve all problems. Evangelists will evangelize, business people will use the right tool for the job, and the vendor-du-jour will be replaced soon enough. It's better to adopt a suite of BI tools that do what they were designed to do best.