Early indicators are very popular in the business world. People normally like them because they see a lot of advantages, such as
- Early indicators are perceived to provide some objectivity to otherwise fuzzy planning and forecasting processes.
- They are most often based on external data, ideally from an institution with a very high reputation. That adds even more trustworthiness to the own predictions and forecasts based on the indicator.
- The external source also serves as an excellent justification, in case something should go wrong. Your forecast was excellent, just the external input data failed.
- Last but not least â€“ there actually are statistically relevant correlations between sets of data which occur with a time delay. If you are able to figure out the right pair of data, the correct interpretation of the earlier events may indeed improve predictions for the later events.
Thus early indicators may have their value for management. However, I am reluctant to rely on early indicators too much. There are some problems that are not easily solved:
- It is very difficult to identify a suitable indicator for a particular forecasting issue. You not only have to find a matching indicator that has some correlation with your businesses performance (or with whatever you want to have indicated). You also have to understand how exactly those two sets of data are interrelated. Does a 50 % increase in the indicator lead to a 50 % increase in your performance? Or will the increase be 20 % / 85 %? Or will your performance actually decrease? Are there any other factors that have an impact on your performance?
- I have seen top managers who demanded â€œan early indicator for our businessâ€ â€“ meaning one indicator. The problem was that this was a billion-Euro business with a lot of product lines. It would have been difficult enough to find one indicator for each business unit.
- Early indicators are easier to find for business activities in later stages of an industries value chain, such as products for end-users (cars, home appliances, services). If you happen to work in an earlier stage of the value chain (e.g. supplier of raw materials or components) you may find it very hard to find something that happens even before you see an impact in your business figures. On the contrary, your results (i.e. orders, earnings, stock quotes) may serve as an early indicator for your customers and for your customersâ€™ customers.
- Early indicators in use for corporate purposes were developed either by scientists with a mathematical / statistical background or by other business people. In the first case, the indicator has more credibility, since it has a scientific basis. However, this scientific basis often is a complex mathematical model that few non-mathematicians will understand. Thus the indicator will always be a bit of a â€˜black boxâ€™. If needed, this fact may serve as a good reason to question the results. In the other case, the indicator lacks that scientific basis. Hence, the indicator could be questioned as â€˜just a best guessâ€™ which is not verified.
- Interrelations between two types of events may change over time. This is even more the case in todayâ€™s dynamic business world. Consequently, early indicators may not be relevant forever. Thus, the indicator should be checked regularly and â€“ if necessary â€“ adjusted. I guess that companies donâ€™t do that all too often.
Hence, I am not a fan of the â€˜One-and-only-Indicator-Approachâ€™. I think that such an approach is not even necessary in most cases. If managers ask for an early indicator, they actually require some sort of consistent data which helps them with their business decisions. Most managers really donâ€™t need something as simple as â€˜If A happens, B follows and we have to do Câ€™. The managers I met know their industryâ€™s mechanisms and drivers perfectly well.
Currently I am developing an information system based on external indicators (still looking for a better name) for my company. I plan to have four groups of indicators â€“ general economic data, our industries data, our costumersâ€™ industries data, and competitor data â€“ and select three to five relevant indicators for each group. These indicators should ideally be available on a monthly basis, at least quarterly. For each of them I will offer a short data history (one year back maximum), a forecast (if available) and a brief comment (as appropriate). There will probably be the possibility for a little drill down to more information. However, basically, thatâ€™s it. I am sure that every good manager who knows his business will be able to make sense of this limited set of data and its development over time.
The idea behind this is not to have one or two early indicators which directly influence business decisions, but to constantly monitor the development those drivers that are most relevant for our business â€“ provided they are measurable.
Note: There is an updated and extended version of this article availabe at our management portal themanager.org:
Early indicators in business – use with care!