Provo River near Bridal Veil Falls from the bike path (click to enlarge)

I spent 90 minutes this afternoon speaking to a vendor of VoIP solutions. I won't give the name to protect the guilty. :-) They had an excellent solution for what I was looking for and a number of features that were really innovative. There was just one problem--a fatal problem if it can't be fixed. I was looking for way to closely integrate a multi-party chat tool with their product. "Oh, we've got a chat agent!" was the reply. Unfortunately, it was woefully inadequate for the application. What's more, chat isn't their core competency. They're never going to resource chat (and a dozen other things) to the point where they can complete across the board. What's the solution? World class integration points.

Companies like the one I was talking to need to pay attention to just two things (at least on the technology side): (a) their core competency and (b) great integration points that are based on standards and easy to use. Otherwise, rather than selling your product's features, you'll constantly find yourself justifying its deficiencies. Much better to say "we integrate with XYZ's chat tool, but I'm sure we can integrate ABC's in a few days" than to try to prove your chat tool (or whatever) can make do to a group of techies who aren't buying your story, or your product.

To their credit, the company I was talking to didn't try to talk us into their chat tool, but rather offered to see what it would take to integrate an outside solution that works. That, however, is an equally difficult position to be in. It joins engineering and sales at the hip and before you know it, all your resources are tied up on small customization projects. I've been there and its not a fun place for a CTO to be. Building integration into the architecture of the product is the only way out.