Why enterprise vendors’ BI midmarket offerings fail

Like clockwork, roughly once a year, the traditional enterprise Business Intelligence vendors announce their newest initiative to target the midmarket. They typically either announce that they’ve formed a new midmarket division, or that they’ve forged new channel partnerships with partners who can bring them into the midmarket, or that they’ve created a new version of their product designed for midmarket customers. The press releases always have quotes about how whatever they’ve announced will make them a powerful force in the midmarket, and will be a great growth opportunity for the vendor. What I find so interesting is the frequency with which these announcements are made. Basically, by re-announcing their midmarket efforts over and over again, they’re letting everyone know that none of their previous efforts turned out as planned.

Why do these midmarket BI offerings fail? It’s not the pricing. It’s not the channel partners. It’s not a lack of focus by the vendor (though this does exacerbate the problem.) It’s the product itself. The vendor always chooses one of two options regarding how to create a midmarket product offering: 1) Dumb-down the enterprise product by cutting out some features that might be considered less useful to the midmarket, or 2) leave the product pretty much as is, create some utilities to make the implementation process a bit easier, and then focus your marketing efforts on touting that this ain’t no dumbed down version — it’s the real deal.

But neither of these approaches works, because at their core these midmarket offerings are still the same enterprise products, with most or all of the same complexity. Midmarket customers rarely have the resources, internal skills, money and time required to manage this complexity and successfully implement, manage and maintain a BI solution. To address the lack of resources and skills, companies bring in consultants to help them design and deploy their BI solution. But, unless you’re willing to sign some type of lifetime employment contract with these consultants, the customer is in for a rude awakening once the consultants leave their premises.

What ends up happening is that the system works for a while after the consultants leave, but as the organization continues to change, the BI implementation needs to be changed as well. Without the specialized resources and skills, the BI implementation eventually becomes a huge mess that looks something like this:

The stack of technology that makes up a BI implementation is pretty complicated. It’s like trying to put a nuclear reactor in your own data center. You could hire some experts to build the reactor for you, but I wouldn’t want to be working in your neighborhood once the consultants hand over the keys (and the nuclear control rods) to you.

The on-demand model solves this problem for BI, which for the first time enables all midmarket companies to have access to a BI solution without requiring specialized resources and skills in house. Just as it makes more sense to have your local power utility company manage a nuclear reactor for you and you just pay for the service every month, it makes more sense to have your BI solution provided as a service. To me, it’s the only viable approach for BI in the midmarket.

posted by Ken Rudin at 2:21 am

---