How Do You Sell BI To The Business Executives?

Whoa! Hold your horses. If this is indeed a key challenge that you’ve tried to address in the past without much success, consider switching jobs. This is not a joke. Business intelligence (BI) is an employee market right now; a key challenge for most BI employers is finding, recruiting, and retaining top — or actually any, for that matter — BI talent. Consider that IBM BAO alone added more than 4,000 (!) BI positions in just over a year! Every other major, midsize, and boutique BI consultancy I talk to is struggling to find BI resources. So if you’ve been fighting this uphill Sisyphean battle for a while, consider new channels for your noble efforts.

Now, some more practical advice — albeit not as exciting. Start from the top down. In a few minutes I am getting ready to talk to yet another large client whose CEO does not “get” BI. Can you rightfully blame him/her? Yes and no. Yes, because how can you manage any business without measurement and insight into your internal and external processes? So if your CEO didn’t learn that in his/her MBA 101, suggest that he/she look for another job. And if you’re still standing after that and have suffered only a mild concussion, consider that many BI projects have been less than successful, and ROI on BI — one of the most expensive enterprise apps — is extremely difficult to show. So can you really blame your CEO?

While there are many successful BI projects out there with tangible ROI (either cost savings/cost avoidance or revenue/profitability increase), they are indeed just that: individual projects, solutions, and applications. There is no business case — or at least none that I am aware of — for “enterprise” BI out there with tangible, proven benefits that disprove reverse causality and eliminate all other variables. Many vendor-hyped white papers are out there, but there’s nothing real, tangible, and repeatable. I found one study that comes close , but it’s very academic and I’ve seen no real-life business application of that approach. However, it is a good read and I highly recommend it for everyone who’s struggling with the basic question posed in this article. If anything, it will show the readers how incomplete, thin, and overhyped most of the vendor-sponsored BI ROI studies are.

Here’s a practical suggestion. Start measuring your BI maturity on a quarterly basis using the Forrester model , compare yourself to your peers and competitors, and then — here’s the key part — start tracking potential correlations between your BI maturity and other corporate performance metrics like revenue growth, profit margins, stock price, and industry ranking. If you do find a correlation, you’ve hit a gold mine and your BI project budgets are safe for the foreseeable future. If you do not, it’s time to stop and reassess your BI strategy. Why isn’t it having an impact on your top and bottom lines? Are you concentrating on the right applications? Are your business stakeholders on board with you and convinced that they are getting real value from BI? Are you addressing key business pain points? Or are you, God forbid, doing BI just for the sake of BI?

If you are successful in changing the minds of your top business executives, the next step is the individual business users. Once again, I am proposing a nonstandard but a very pragmatic approach. Unfortunately, BI is not a grassroots environment. One can’t tell a business user: “Stop using Excel, start using our new enterprise BI apps, and your life will become rosy and wonderful.” No, no, no! That’ll be the end of your good relationship with your business counterparts and they will forever label you and your efforts as an “IT nuisance.” Admit that their life will become more difficult, but it’s got to be done for the sake of a single version of the truth, better collaborative decision making, etc, etc, etc. But you have to give these people something in return, as we human beings will probably never do something “just because” someone says so. It goes against our human nature. So everyone has to have some individual, personal incentive. And what can be more effective as an incentive than tying the usage of enterprise BI to individual performance reviews, which could affect BI users’ variable/incentive compensation? My apologies for being this blunt, but I have indeed seen multiple cases where this works extremely well.

In order to implement this approach, you need BI on your BI environment to understand who uses (or doesn’t use) your BI applications — and when and how. You can get this data from most of your BI tools and/or databases. If you have multiple tools, consider Appfluent or Teleran, which will provide you these statistics on multiple BI tools and multiple DBMSes. If much of your BI environment is locked in spreadsheets or file-based desktop applications like Microsoft Access, you can monitor their usage using tools from vendors like CIMCON Software, ClusterSeven, Finsbury Solutions, Lyquidity, and Prodiance. If it sounds like BI on BI may become a project of its own — well, you’re right, it certainly will. We strongly advocate that BI on BI become a function of your BI center of excellence (COE) or competency center (CC).


There is another way to prove value

As usual, I completely agree with Boris' main point here, but one other way to quickly demonstrate the value of BI is to find ways to use your data to enhance the products that you sell. There are many examples out there. Brokerage houses analyze the performance of client portfolios. Media companies analyze the performance of purchased placements and make recommendations to customers. Package your valuable data and sell it as a service. Often, I find BI is so busy looking inward for value, it misses opportunities to provide value to customers and partners.

Lead by example

I have found that the best way to make the BI argument is to build a prototype solving a critical business problem. Once you can reduce uncertainty or improve productivity or enhance decision making, you will get people's attention. Start small, figure out how to address the business problem quickly without a long development cycle, keep the cost low and before you know it people will be coming to you instead of you going to them. Make sure you pick a high visibility business issue. Partner with someone from the business. Keep the expectations low and the buzz quiet until you have something to show and then lead by example!

Embrace spreadsheets - don't compete

A very good article but I would take issue with the implicit tone that spreadsheets are the 'enemy' of good BI.

The truth is that spreadsheets, Access databases, CSVs etc are PART of an integrated information management environment, just as cars are part of an integrated transportation system. Planes and trains can service mainstream needs but every individual needs the flexiblity to 'do their own thing'.

Like it, or not, almost all corporate financial innovation takes place in spreadsheets. This innovation will be on a road to production or redundancy depending on its continuing value. The conversion from spreadsheet to system usually takes many months if not years. By the time a valuable spreadsheet is replaced by system enhancements, new business needs will have created new spreadsheet replacement is a continuing journey not a destination.

Once it is recognised that spreadsheets are the 'yin' to the 'yang' of BI systems then one can embrace this unstructured financial data source (using the right additional technology) to great value ....just as Google and Autonomy have shown in the world of text etc. Not only will this transparency show where IT investment should be properly targeted to benefit the business you will also truly have 'BI on BI'.

Not sure how you interpreted

Not sure how you interpreted my tone. If you follow my research and my blogs you'd see just the opposite - Excel is a #1 BI tool and it's here to stay. I always recommend embracing that approach (rather than fighting it), but managing it appropriately.

...probably being too sensitive

From our past conversations I know you are very pragmatic on the issue of spreadsheets - but I still think the article suggests that the choice is BI System OR Spreadsheets rather than BI AND spreadsheets.

Once one overcomes this hurdle I think that the really valuable missing dimension of the conversation is how to properly combine BI with spreadsheets in a business. That conversation must recognise that people want to do a lot more than just view data slices.

Most BI vendors pay lip service to the Excel dimension - they are obligated by client demand to have an 'export to Excel' button. However when you sit through their demonstrations this is usually seen as the 'ejector seat' option rather than creating a continuum between system and spreadsheet.