Well, the marketing department has requested, and you got the job. “Build us something that taps into all that social data out there and lets us sell more. By next Tuesday, eh?”

What are you going to design into such a system? What’s its core idea?

If the idea is “sell more product” as a direct consequence, you’re going to have a system no one is happy with — in fact, it may even backfire.

Ah, but if you make its core teasing out who’s a useful influencer for your firm, that’s a very different story.

I found out yesterday about a not-for-profit in Ontario that builds out solar power generation facilities. It has the power contract in place, with prices guaranteed; individuals can buy five year bonds that pay a decent return based on that contract, and the capital raised builds more solar generation capacity.

Unbeknownst to this little not-for-profit, I’ve written three different articles — one that goes into national distribution in the business pages of the country’s newspapers, one on my own blog, and a third for a specific requirement — that have told this organization’s story. (And here I am doing it again.)

Suppose you could put things in front of someone like me that I would respond to, that would increase awareness of your organization, and opinions about it.

After all, no one other than a government body has more than 50% market share, usually. Mostly, they don’t even know you exist.

People get in ruts. Those that have never bought anything other than Green Giant or the Loblaws house brands (President’s Choice, No-Name) don’t even notice the Le Sieur on the shelf, or know about Selection, Irresistibles or Compliments.

Influencers wake people up and cause them to look around, or look for you. Now you can sell something.

What this little exercise is about is pointing out that the requirements around many applications that are designed to chew through data to find things will be inappropriate to the desired outcome.

It’s going to be up to us to make sure we give our enterprises what they need, and not just fulfil the stated wants. Otherwise, we’ll kill the trust and expectation that information is useful.

Do that, and IT in-house has no future.