What Market Share Matters?

Both Lessien and John Gruber take on the topic of market share and I think they miss some of the nuances.  

Lessien applies the basic business school approach: 

Large market share attracted developers who built software exclusively for the dominant platform. That software, in turn, created further lock-in as users grew accustomed to the workflows and proprietary data formats that emerged. Typified by Microsoft’s “embrace and extend” strategy, market leadership yielded a nearly permanent advantage, which suffocated competing platforms and deprived customers of choice. Essentially, the historical advantage of dominant market share has been the ability to raise (discriminately) the switching cost of competing platforms.

In other words, there are massive network effects in technology.  These network effects can lead to monopoly rents.   

While she doesn’t do so explicitly, Lessien goes on to suggest these network effects aren’t applicable to mobile. Despite even a commandeering market share, monopoly rents can’t and won’t be created and therefore market share is largely irrelevant.       

Gruber takes a slightly different tack by suggesting market share and profitability are loosely correlated, but that this correlation has been minimized in the world of mobile. In paraphrasing Lessien, Gruber states, “profit share seems a better indicator of success than market share — both today, and historically.”

I think the three of us agree that the most desired applications from a user perspective, the “table stakes applications” that represent the top 85-90 percent of desired applications will be ubiquitous to all platforms. I think we also agree that the perceived horse race created by pundits shouting every month when the market share metrics drop is overdone.  Everyone wants to catch the inflection points, but these inflection points will never materialize as a single number.

But let’s take some of this to its limit.  Developers are constrained.  Big players do have the ability to ensure their services are ubiquitous to all platforms and thanks to web applications the  “table stakes applications” are (or will be) available on the leading mobile platforms. But with what lag? Even a short lag creates network effects. Consumers, knowing that their horse will always finish, but never first (sport analogy for Gruber) will be inclined to change their bet to a platform that gets the newest and next table stakes applications first.

The question that should be asked is not “does market share matter?”  The question that everyone should be asking (but no one seems to be asking) is “what market share matters?”  Ease of monetization is one to consider. Gruber indirectly hits on this: “[Apple has] been behind at least one giant (Symbian at first, now Android), and for most of that time they were behind BlackBerry. iOS’s influence on the market has always been disproportionate to the iPhone’s share of the smartphone market.” While Lessien thinks Apple’s leadership position will dissipate, I’ll take the over on how long that takes. Regardless, being the market share leader in app sales clearly creates network effects that are captured by the hardware vendor (not to mention the 30 percent rake on the app purchase price).    

The mistake market share gawkers make and the one Lessien and Gruber fall prey to is working from a western viewpoint where profitability is defined at the point of purchase.  Multiple some margin level by the number of units sold and viola. But this isn’t necessarily how market share drives profitability.

For example, until recently market share was a defining property of profitability for many technology categories like televisions because it was a good proxy for what was happening behind the scenes and not because of what happened at the point of sale.  Many Japan-based companies for example are frequently chided for their heavy focus on market share but many of these critics fail to realize how this focus has historically translated into profits.  Historically Japan OEMS had entire ownership of their production facilities.  Joint ventures are a relatively new phenomenon among these corporate citizens and they’ve only recently moved to China. In the early days of television for example, the real money was made by keeping capacity utilization of underlying factory high.  These companies made their money on the turn not on the sale of the underlying item.  What the television sold for – even if at a loss- was secondary.  Some of this is what naturally drove television prices lower over ensuing decades. Companies – or rather factories – competed with each other to keep their utilization rates high.        

Lessien and Gruber are right to point out that market share of unit sales isn’t what everyone thinks it is. But rather than dismiss market share in its entirety, we need to look at what market share matters. I agree in the case of mobile, it isn’t unit volume. I’m not so quick to dismiss network effects in mobile, but do agree they aren’t driven purely by unit volume market share.