Sunday 21 October 2012

Anchoring & Estimates

I was in a meeting recently discussing a sale between two software systems and a potential large client. Integration between the two systems came up. It was a high level discussion as the project is not funded yet. Even though it was still at 'Pre-Sales' stage, the client was looking for an estimate on how much effort (and cost) would be involved in the integration aspect.

Without substantial analysis, it is hard to give an estimate and there is no time or funding to do the analysis, so you have a bit of a  chicken and an egg. A sale being a sale, both software companies did not want to over complicate the situation and embark on time consuming pre-project work. If they did, the deal would in all likelihood fall through.

One software company said they thought it was about 3 days work for them, the second company then came up with 5 days as their estimate.

I was talking afterwards to the person who put forward the 5 day estimate and asked him how he came up with that. "Well when the other guys were doing 3 days and we have to do a little bit more, 5 seemed reasonable, it would have created doubt if we said we don’t know".

This is classic Anchoring Bias, one of several cognitive biases that have been covered in many studies. Psychologists describe it as a human tendency to rely too heavily, or "anchor," on one piece of information when making decisions. An example of this in a business context would be an investor judging a stock price as overvalued or undervalued based on the stock's previous high price. The previous high price is the anchor. It is part of the larger study of judgement and decision making under uncertainty which is best exemplified by the work of  Daniel Kahneman and Amos Tversky.

The scenario above happens all the time, be it people being asked to give sales estimates, costs of production, number of people needed, or time lines for a project conclusion.

Where well analysed, firm evidence is lacking or suspect (as is often the case) then someone usually gives a ‘Guesstimate’ and this becomes the figure around which all other estimates are given. The origin of the guesstimate is often how long a similar task took or costed previously. Bold and convincing arguments are then needed to dislodge the guesstimate. Think back to a few meetings you have been in and you will recognise this.

So what can this mean for you in your work place? Well, one implication is that if you are going into a meeting where you know estimates will be required, get your estimate in first and let that dictate the pace, otherwise you will have to dislodge someone else’s anchor and that is easier said than done. If you understand the sports cliché 'get your retaliation in first'  you get the idea.

If you want to deliberately control the anchoring of any figures, then look at the agenda, speaking order, seating order, content of your presentation etc as a medium for setting out your numbers. If you can't prevent a number being anchored by some one else, be ready with a suggestion to have it bottomed out in another meeting or some other process. Letting it happen by accident is not good enough.

Getting back to my opening account, the second software company will now find it very difficult to charge for more than 5 days work, will really have to justify it if they do .That is not a position a project manager wants to be in.

No comments:

Post a Comment