Twenty billion flies can't be wrong...

You can't afFord not to...

One of the more obvious beneficiaries of market share statistics was the Ford motor company. For decades every company fleet was predominantly Ford, due to deep discounting by the manufacturers (sometimes to the extent of selling cars for less than they cost to produce - although this was not unusual at some times in the history of the motor industry). The result was a saturation of Fords on the market, leading to the largest dealer network - which in turn meant that private buyers would buy Fords because there was always a local dealership.

It was common knowledge that the cars were not the best. With some notable exceptions (the Mark I Escort, for example), Fords were mechanically unremarkable, poorly equipped and often badly finished. It didn't matter: as long as they lasted three years in fleet use, they had done their job. After that, the sooner they rusted and fell apart the better.

The Volvo 340 was pretty good in its day. You could buy a 340 with a 1.4 litre engine, equipped with electronic ignition, side impact bars and a part-galvanised body, for less than the price of the similar-looking Mark III Escort 1.3. The Escort was the one which sold. No matter that the Escort was unlikely to push the limits of ifs five-digit odometer before mechanical disaster struck: there was always a Ford dealer nearby. Because, of course, there needed to be.

There is no doubt that in the early 1990s almost every model in the Ford range was comprehensively outshone by its General Motors rival, but still Ford ruled the roost in the fleet market, which in turn gave private buyers a feeling of comfort and safety. It is very noticeable that now Ford have lost much of their core market to better-built German rivals, they have suddenly started making reasonable cars. Whether they could have done so all along, but chose not to in order to save money, is a point on which it is interesting to speculate.

How to lie with Statistics

As with most statistics, research is carefully tailored to provide the required result. For example, one well-known software vendor published a survey of forty-one carefully-selected Fortune 500 companies, and extrapolated the results to imply that their email software was market leader. Independent research from IDC, with a much wider and more diverse base, showed that this was patently false². But that didn't stop the advertising campaign: a lie, if told often enough, becomes the truth³. The result was that the real market leaders⁴ are left playing catch-up in the press war. And since the firm with the biggest advertising budget gets the most coverage, they were on a hiding to nothing⁵.

A bundle of lies

Market share data is also vulnerable to manipulation through licensing systems and "bundling"⁶. Every single Windows PC comes with a copy of Internet Explorer and Outlook Express, making these by default the most common web browser and email client. Windows XP includes an Exchange client license, meaning that every copy of XP sold is counted as an Exchange "seat" even if this license is never used.

People use various email clients in an attempt to avoid the constant stream of viruses which exploit Outlook's non-existent security, but these users are still counted as Outlook "seats" because they have a license for the product. Market share data for Outlook is therefore inevitably and fatally distorted by the fact that it can't distinguish between active and dormant licenses.

And the same goes for Internet Information Server (IIS), a product so insecure that Gartner Group recommends existing users to switch away. Indeed, the majority of webmasters don't use it even though it's free. No matter: they still have a license and appear on the market share data, even though research shows that Apache is the market leader in live web servers.

Share And Enjoy

This can lead to some extraordinary results. During the mid 1990s, Ford sold huge numbers of Escort cars throughout Europe. The Escort was the biggest-selling car in many countries. Yet anyone who has driven one of the oval-grille Escorts will tell you that it was probably the worst car on sale by a serious manufacturer at the time. The engines were noisy and harsh, the suspension was woeful, they wouldn't steer in a straight line, the trim rattled, and they were in every measurable way worse than the model they replaced. But "you can't go wrong with an Escort," "look at the number sold," "we're selling millions" and all the other spurious arguments of market share were brought to bear, pushing millions of these awful objects out of dealerships and onto the nation's roads. The BBC's Quentin Willson repeatedly panned these vehicles - as he pointed out, for the same money you could buy a better car from almost any other manufacturer.

And recent advertisements for digital TV providers show a new extension of the game. "We carry more football than they do!" comes the cry. And of course they must do, since they have licensed all the competitor's channels and provide one of their own. What will happen to the quality of service if everyone transfers to the supplier with "more" channels thus reducing the profitability of the other supplier? Will the "more" supplier take over content production and lay themselves open to the same attack? Perhaps not.

Fear, Uncertainty and Doubt

A key purpose of market share statistics is as a tool to spread FUD⁷. Thus, "they are tiny, we are big, their product is doomed, ours will soon be as good, and if you invest in their product you'l be wasting your money."

This is particularly important when markets are in a state of flux, when new technologies are emerging. Owners of Betamax video systems suffered exactly this problem: their equipment was technically superior, and indeed is still used where quality of signal is an issue. They had automatic tracking control, a far less complex tape path (meaning less wear in the components and tapes). What do we all have now? VHS. The marketing machine used FUD to persuade those releasing pre-recorded tapes that the Betamax market was small and shrinking, so the film companies concentrated on VHS. Which made prices for Betamax higher, and locked into a positive feedback cycle which quickly led to Betamax imploding.

Another example of this triumph of market share over sense is in computer operating systems. In 1984 Apple launched the Macintosh, the first mass-market computer with a graphical interface. Microsoft, whose MS-DOS software was the market leader for personal ocmputers, were seriously worried. A series of graphical interface products for PC became available, of which GEM was one - but they relied on underlying MS-DOS or compatible operating systems. Microsoft began a FUD war: DOS was not guaranteed to support these packages, but it would support windows. Some versions of DOS were released which cuased competitors to the emergent Windows to crash.

Early versions of Windows were buggy and feature-poor by comparison with their competitors, but sufficent FUD was spread to give Microsoft time to release Windows 3, the first reasonably stable release. The copmpetitors sank without trace.

By the mid 1990s Windows had the dominant market share, to the extent that Microsoft's biggest competitor in 1999 was pirate copies of its own software⁸. In 1995, though, Windows was stagnating: it lacked the plug-and-play and built-in networking features which the Macintosh boasted almost from day 1. Apple release System 7, then System 7.5, which improved multi-tasking and began to compete strongly with Windows. Microsoft's response was Windows 95 - which in its first releases was buggy, unstable, and didn't support Internet technologies (Microsoft tried to get you to use their proprietary Microsoft Network instead). Once again the marketing machine used FUD and market share again to avoid losing market share while they "got" the Internet.

The mantra was repeated in all PC publications: Apple are doomed, they will be dead and gone any day, Windows is the future, Apple's market share is too small and people won't write software for Mac... and on and on. Now, we all know that Mercedes-Benz and BMW combined have a smaller market share than Apple. Are they doomed? They like to think not. But there is an element of fear that we are locking into the wrong technology. We doubt. Apple were the largest supplier of computers in the education market in the US, almost universal in the creative arts, even Bill Gates had one - but the perception was created that they had a vanishingly small niche market.

Being By Way Of A Conclusion

The moral of this tale is: if you're not a fly, don't eat the manure.