Examining the situation, I found that the outstanding amounts almost all arose from one of three reasons:
- Software didn't perform in accordance with documentation - "the OMGEO gadget doesn't interface properly with the FX widget";
- Professional services had been poorly performed - "Yes Fred was on site for 20 days but he only did 10 days useful work and that is all I will pay for"
- Maintenance charges were unclear - "I can't reconciile the invoice you sent with the five contracts we have."
None of these could be resolved by finance staff; they needed the technical knowledge of development, professional services and legal respectively.
Once this had been recognised, and the finance staff given permission to call on others for help in resolving issues, the problem went away and did not recur.
In the primitive world, this was a useful, practical and economic way of working out what was wrong. In a world of complex tasks and large organisations, it often leads us to wrong conclusions.
There are several ways of overcoming this bias:
- Look at what the majority of people do in the situation (in the example, we had about ten people in five different locations, all performing poorly).
- Ask what you yourself would do in that situation, given the structure of the organisation, the incentives and the information available to you (in the example, the finance people did not have permission to insist on others helping them to resolve prolems);
- Look for less-obvious causes.