Today I’ve watched Greg Young’s keynote from DDD eXchange 2016. The talk is really awesome, though-provoking, turning everything upside down… I really recommend.

However, the point that caught my attention most was the answer to the question from the audience (at the very end). Greg mentioned that we need to be very careful with who we consider domain experts. He even said that on one project their end-users hated the app, because their “domain experts” actually had no idea about the work that end-users were doing.

I’ve experienced that mistake myself, in one of my summer jobs. For a couple months I worked in a call-centre for a mobile operator.

When you call somebody in the middle of the day to talk about their soon-expiring-contract with a mobile operator, what do you think is the most common answer?

I need to think about it/I can’t talk right now, please call me at XYZ time.

Pretty obvious, right?

The only problem was that it was impossible to do, because the calls were assigned randomly to people and it was impossible to make sure the call was assigned at a specified time! You could specify the preferred time range, but more often than not it didn’t work as you’d expect.

The result?

People entered random information and date in the system, wrote everything down in their paper calendars and handled the whole case outside the system. Some time later another unlucky operator who got that call assigned had a very bad day (you won’t believe how angry people are when you have no idea what they told your colleage a few months earlier and what they asked them to do back then).

In the end the system was an expensive phone book, that pissed off a lot of people, both working there and those that got the calls. The work was hard, stressful and annoying enough in itself, and the software we had to use only made it much worse instead of helping.

That’s what happens when you don’t realize you’re dealing with proxy domain experts.