I finally got round to watching this video and it was worth it. I don't like the slightly cult-like way that the message is presented or the I-am-totally-right and everyone else is totally wrong way in which John Seddon presents his ideas, but it is always good to hear a challenge to orthodoxy and the basis of his critique is interesting and probably has a lot of validity. Here are the key messages as they struck me:

  • its a mistake to try to make a clear split between the front office and the back office with the former focussed on the messy business of handling customers and the latter aiming to become a super-efficient processing machine;
  • a clear split is impossible because in fact lots of stuff will keep moving back and forth between the front and back office and this can't be eliminated;
  • morale will be low in both parts of the organisations because neither group will have much job satisfaction;
  • customers will be dissatisfied because they will be passed around or left dealing with a disgruntled and impotent front office;
  • the whole thing won't really be efficient because the process flows for a problems will typically be long rather than short.
So what is the alternative? This isn't set out too clearly (and it seems to me almost inevitably that one will have something like a front office and a back office even if one does not aim to split them off from each other as clearly as possible). To a certain extent it would seem to be just the opposite of the above. But here are a few more pointers (all based of course on my interpretation of what I heard):
  • focussing on understanding what the customer wants and then designing systems (or processes?) that satisfy this demand;
  • try to solve the problem as near to the customer as possible - this is more satisfying for the customer and for the staff;
  • be very careful about targets and management information since this very easily drives non-optimising behaviour;
  • targets should be broad, outcome-related and owned by the teams whose work they relate to, so less "% of calls answered with X rings" and more "% of customer requests solved by the team"?
Anyway, as I said, good to see someone challenging conventional wisdom!

posted over 2 years ago