Saturday, 2 March 2013

It's the stupid data

When I was first looking for a job my father's advice to me was to join a company with a good reputation and brand name in a market that makes good profit margins.  Pretty good advice, which for the most part I have ignored.

Instead I have found my self working for and running businesses that operate at wafer thin margins and its a hard grind.  The job of a manager in these types of business is two fold: firstly find a way of navigating to markets with improved margins and secondly to run a very tight ship.  To run any business where the profit to turnover ratio is under 7% requires one thing; a trusted set of numbers for managers to work from.


Managing the economy is just a complicated version of running a low margin business (without the opportunity of navigating to high margin markets) and its becoming clear that governments and their civil servants are trying to do the job with one hand tided behind their backs.  Specifically in the UK our Organisation of National Statistics (http://www.ons.gov.uk/ons/index.html) is in disarray.  An example of this is the endless revision of of GDP numbers up and down, begging the questions - are we in recession, were we in recession and are we likely to fall back into recession?  Imagine trying to run a company when you have know idea of forward monthly cashflows?  Similarly the ONS measures inflation in two ways, one is known to be faulty (RPI) but they continue to waste resources on collating this metric.

Is that the north star or a satellite?
The job of collecting the numbers is becoming more complex with the advent of computerised trading, e'commerce and the globalisation of manufacturing.  This last point means a car made (finished) in the UK may in fact contain only a small proportion of UK 'value add' with most of the valuable components being imported before assembly. All this complexity begs the questions; what should we measure and how can we be more accurate.  The same issue applies in a company where the contents of the 'Board pack' needs to inform the decision making.  Having designed Board packs for directors in a number of companies I have come up with a number of rules:

  1. Only measure stuff that informs a decision the Board will make - less is more
  2. Focus energy and resources on management information that helps drive good decisions that relate directly to the Boards objectives - other stuff doesn't interest them
  3. Leave historic reporting to the Auditors - you can't affect what's already happened
  4. Stop measuring things that you can't measure accurately.

The onus for improvement is not with the ONS it rest squarely with the Executive (the Cabinet) and there are a few things that would make a difference.  Firstly, they should revisit their objectives and try set objectives that they can determine, there's no point having a set of objectives that will be largely governed by external global events. Having clarified their objectives they should be clear about the data they will use to measure progress towards these goals. They should then focus the ONS's resources on this data, switching resources from what is interesting to what is going to be used.

Blog Archive

Subscribe Now: Feed Icon