One of the great spectator sports in the UK has to be the Antiques Roadshow. You know what I’m talking about. It’s that moment when the expert appraiser turns towards the person who has just lugged great uncle Herbert’s ugly little chest of draws across town and tells them that its worth a small fortune - often too much gasping and big smiles.
Most of the time you can tell what’s coming by the comments, but occasionally there is a surprise. Forget unknown unknowns - this is the ‘we don’t know what we have until someone else tells us’ and has done more than anything else to promote the car boot sale. So what you may ask has this got to do with forecasting?
Launching a new product onto the market place is a costly business, what with the development work, the pack design, the product design, the push to get it on shelf, setting up new manufacturing processes and negotiating contracts to buy new materials - It all adds up.
Richard Kottler maintains that the smart companies know a thing or two and he reckons that the successful companies are successful behind two key competencies: strategic marketing and market research.
But even the smart companies struggle when it comes to running a volume forecast, often balking at the cost of such an exercise. Why spend so much on something that is by definition wrong?
Here’s a real account of how one company tackled the issue and how they came to view forecasting in a completely different light. The big challenge facing their market research department was that they wanted to run forecasts, but were facing strong internal opposition due to the poor record of previous forecasts.
Thinking that they deserved more credit, they set about reviewing their current processes, collecting all the information they could find about previous projects.
This was not as simple as it sounded. Companies tend to be great at generating data but very few have the discipline or are good at creating useful databases (a few numbers in an Excel spreadsheet doesn’t count).
Anyway, the company (or rather a few geeky fanatics) toiled away collecting whatever data they could - what forecasts had been run, what data had been used, what decision had been taken, had forecasts on launched products been updated and were they tracked in market. How accurate were the forecasts?
This was a huge exercise, not only because there was a lot of data, but also because it was so hard to find and getting people to help collect data from yesterdays new product was not sexy or exciting. Finally, when all the work was done and the results had been created, they were able to sit back and survey their creation, ‘are we any good at forecasting?’
Err... No, not really.
End of the story you think? Well it might have been except for the hero of our story. Now if you are expecting the hero to come from the market research department, well sorry - wrong. And no, it wasn’t a brand person either. It was actually someone from the finance department who, like great uncle Herbert’s chest, inherited a goldmine, except he just didn’t realise it quite then.
As he looked at the mass of information he had inherited, he gravitated to what he knew - the key financial data.
As he looked at the data for the new product launches he noted something rather interesting, which set him looking for other key financial data.
When he laid this data side by side, he saw that yes, the company was not actually very good at forecasting, but also that they tended to be not very good in one particular direction.
In fact many of the forecasts tended to be quite optimistic. Consequently, he also noted that in the cases when the forecast was too high, the initiative tended to miss its financial target by much more than when the forecast was lower.
He also noted that when forecasts were within a specific range, the initiative tended to hit its financial targets 100% of the time.
This might not be so remarkable had he stopped there, but he went on to ask two questions: How good are we at forecasting? And, can we do better?
What he found was that the company was much worse at forecasting than many of their competitors. This then led him to rephrase his second question - instead of asking can we do better (the answer to this was obvious) he asked, why would we want to do better?
A little further digging allowed him to argue that if the company could improve the quality of its forecasts by a certain percentage, this would have a huge impact on the financials related to any new initiative.
And if this was rolled out across the whole company, the impact of this small improvement would be a massive contribution to the company’s bank balance.
This led to some real attention and focus on improving the company attitude towards volume forecasting, shifting it out of the realm of a pointless exercise to something which was able to make a solid contribution to the bottom line of the company.
Forecasting, just like great uncle Herbert’s chest, was suddenly looked at in a completely different light and had pride of place in the corporate living room.
Now, I am not advocating you go running off to the nearest car boot sale because it’s quite unlikely that you’ll end up with a priceless work of art.
What I am advocating is that thinking about forecasting in the right way can certainly help to ensure that the company treasure chest is an heirloom you can bank on for the future.