Estimates and Anchors

Sometimes you will be asked a question and you won’t know the answer…

  • How many customers will call?
  • How many products will you sell?
  • How much will the project cost?

Sometimes you have to start with an estimate.

How do we estimate?

Let me explain with a question: how many people live in my home city, Nottingham?

When we estimate we always start with what we know.

The American perspective

If you are reading this sitting in California (I have readers in California, why they are reading this when they could be surfing is beyond me) you will perhaps know that…

  • Nottingham is a provincial city in England
  • Robin Hood once lived here (So did Kevin Costner, Prince of Thieves)
  • It is smaller than London
  • The population of London is about eight million people

So what is your guess? Smaller than eight million, but big enough for me to have heard of it, let’s say…

One million people.

The British perspective

If you are reading this sitting in Mansfield (an English town and somewhere that if you are from California I recommend you avoid at all costs, particularly if you like surfing) then you will know that…

  • Mansfield is the second biggest town in Nottinghamshire (after Nottingham)
  • The shopping is a whole lot better in Nottingham (we have a branch of Waitrose, not that I am a snob, just socially superior)
  • The population of Mansfield is about one hundred thousand people

Now what is your guess? Bigger than one hundred thousand, better shops, let’s say…

Two hundred thousand people.

The power of the anchor

The point of this little exercise is to show that your estimate is always anchored to your starting point, so your starting point introduces a bias.

  • If your point of reference is London, then chances are you overestimated the size of Nottingham
  • If your point of reference is Mansfield, then chances are you underestimated the size of Nottingham

The answer (which depends on your definition of Nottingham and there is a whole other debate) is about six hundred thousand people — six hundred thousand and one when Kevin is in town.

So how could you improve your estimates?

If you are going to estimate something — and sooner or later you will — you should always write down your assumptions, and get somebody to sense check them.

And use at least two anchors

Particularly if you are a Californian who is estimating how far he can surf off Nottingham’s golden sands.

If you enjoyed this post click here for updates delivered straight to your inbox

Nottingham Riviera

Read another opinion

Image by Simon Collison

Comments

  1. Hi James — My younger sister lived in Nottinghamshire for several years when her husband was on assignment with Capital One. Rather than estimate the size of Nottingham, I probably would have called her for a more precise reading. Which gets to the point of tapping a primary resource whenever possible (“phoning a friend,” or a sister, if you will). I guess that puts the anchor in just the right spot.

    I tried to dig up her old address to see if you might have been neighbors, but Nottinghamshire was the best I could find. Regardless, it is definitely a “small world” (but I’d hate to paint it, as comedian Steven Wright always says).

  2. Cathy Field says:

    James, yet another insightful and funny post. You manage to always keep topics interesting. Thanks for the laugh and learning rolled into one nice bite.
    I lived in Nottingham for a couple of years and I guessed wrong!

  3. I recommend you read Doug Hubbard’s book, “How to Measure Anything,” with special attention paid to Calibrated Estimating.

  4. James,
    One thing that I learned as an economist (in a previous life), was that the forecast or estimate could never be wrong as it was based on assumptions. Therefore, if we want to improve our estimates or forecasts we need to get better at the assumptions that we make.

    By the way, the council has the population of Nottingham as just over 300,000.

    Adrian

    • James Lawther says:

      Thanks for your comment Adrian,

      Your 300,000 comment is interesting. Nottingham city is one of the tightest council boundaries in the country. It only encompasses the inner city. So if you live in one of Nottingham’s more affluent suburbs then chances are you don’t actually live in Nottingham. you live in Broxtowe or Gedling or Rushcliffe, all of which are much more desirable.

      Which explains why the city is often referred to as one of the poorest in the country.

      You are very right about assumptions and you also highlight an interesting point about definition

      James

  5. One of the best books I have read on this subject is “How to Measure Anything” by Douglas W Hubbard. I recently wrote some blogs linked to his methods as part of a review I was doing of the book. Some of it is fairly easy to follow other parts of it demand a re-read and concentration! It is an enlightening text and I would recommend it to to those who have to wrestle with estimating the values of so called intangibles.

    You could start here with , I hope, my simple take on it before tackling the book itself.

    http://theifm.co.uk/book-review/can-measure-anything

  6. Emine Gokce Phillips says:

    Hi James, a very good read and a very interesting topic indeed. How about a mathematical/statistical approach to estimates or trends?

    For instance, what if the question in your article was modified as follows:
    – Nothingham had 250K residents in 1940, 210K in 1950, 300K in 1960, 360K in 1970…. 500K in 2000 ( all made up numbers for the sake of the example)

    What kind of tools would you/others following this group use to estimate its future population – say in 2050?

    Would a simple Minitab Trend analysis suffice (or at least give a good indication about the ‘to be’ state)?

    How would you include the factors such as war, floods, economic crisis, etc?

    • James Lawther says:

      Emine,

      That is a good question, I suppose it comes down to hard work (cost) versus the implications of getting it right (benefit).

      I guess there is a way of risk modelling that, but I’d be lying if I said I knew what it was

      James

  7. Hello James

    I find anchors fascinating: they occur early in life and continue thereafter. You were anchored in the English language in such a way that you do not even know you are anchored. You have been anchored with the English way of life and you are not really present to that anchoring until/unless you come across someone who acts very differently – bizarrely.

    Of all the anchors, it occurs to me that the most damaging ones are:

    – anchored to being small and inability to affect the bigger game of life;
    – anchored to staying safe not putting the head above the crowd and thus being open to being ridiculed;
    – anchored to obeying authority figures and finding-filling your place in the hierarchy;
    – anchored to finding the one right (correct) answer;
    – anchored to taking others views as truth and making them your own …..

    It occurs to me when you get present to this anchoring the other anchors like population size as pretty trivial types of anchors. What do you say?

    All the best
    maz

Speak Your Mind

*