We spent some time this morning with the man with the hardest job in the world, Yusuf Mehdi of Microsoft. Why does Yusuf have the hardest job in the world? Because he’s responsible for building Microsoft’s Bing into a real business. Yusuf was charming and forthright, as always, and he shared some details on Bing’s progress over the past year. As we’ve noted, this progress has been significantly more impressive than we thought it would be when Bing launched last year.
That said, having walked through the numbers again, we remain convinced that Bing will never become a good business for Microsoft.
In short, we just don’t think Bing will ever be able to gain enough market share against Google to achieve the scale necessary to become significantly profitable.
LET’S GO TO THE NUMBERS
Bing’s first problem is that, per Yusuf, Bing’s cost structure has been built to handle 40%-50% of the U.S. search market. (These costs are primarily server and bandwidth costs–it costs a lot to aggressively index and store the world-wide-web). This explains why Microsoft’s Internet division is currently losing about $2 billion a year.
To turn profitable on the existing cost structure, therefore, Bing has to capture 40%-50% of the U.S. search market, up from the 12% it has now. The Yahoo deal will help, but not nearly as much as most people think.
Here’s the way the US search market breaks down right now, as measured by approximate query share:
Ask Network: 3.5%
Another way of looking at the share is this:
Even if Bing were to capture 100% of the non-Google U.S. search market right now, in other words, Bing would have less than 35% of the market. Just as important, Bing’s “owned-and-operated” share of the market would only be a measly 12%.
This highlights Bing’s second problem: There is a huge difference between the profitability of “owned-and-operated” search queries–queries conducted directly on Bing–and “partner” search queries: Results served by Bing of queries conducted on a partner site, such as Yahoo.When a query is entered on Bing, Microsoft keeps 100% of the revenue. When a query is entered on Yahoo, Bing will only keep 10% of the revenue (Yahoo gets the rest). Contrast this with Google, which keeps 100% of the revenue of a query on Google.com.
Right now, only 12% of U.S. search queries are conducted directly through Bing. The additional 17% or so of market share that Bing will gain when it concludes the Yahoo deal will not be owned-and-operated. This means that Bing will only get 10% of the revenue from these queries, or the equivalent of another 1.7% of share points on an owned-and-operated site.
To gain enough market share to turn profitable on its current cost structure, therefore, Bing must do the following:
- Capture all the share of the market that Google doesn’t currently have (35%, up from the 29% the company will have when the Yahoo deal closes). This means stopping Google’s market-share gains in their tracks.
- Convert at least 10 points of market share from “partner” site queries to “owned and operated” Bing queries. (Remember, on partner sites, Microsoft only gets 1/10th of the revenue, so it needs 10 points of partner site query share to get 1 point of owned-and-operated revenue share). This will mean continuing to steal share from its “partner” Yahoo.
- Steal at least 10 points of market share directly from Google (So far, Bing–or anyone else–hasn’t gained any share from Google.)
The last challenge in this list, in our opinion–stealing 10 points of share directly from Google–will be the most difficult. Google is not run by idiots. We’re impressed with many of the innovations Bing has rolled out in the past year, and, as consumers, we’re glad that Bing is shaming Google into implementing many of them. But we continue to think Google will copy any interesting innovations that Bing rolls out — and therefore neutralize Bing’s attempts to gain significant market share.
These moves would give Bing about 45% overall share of the market and 30% owned-and-operated share. We don’t think Bing would be a wildly profitable search business on this share, but it would probably be nicely profitable. That said, we consider it highly unlikely that Bing will ever gain this much share, especially the 30% on the owned-and-operated Bing.com.
WHY NOT JUST CUT COSTS?
Can’t Bing just cut its cost structure? What’s wrong with making a nice living on, say, 25% of the search market?
That’s a nice idea, but the problem with the search market is that there are major economies of scale. The more queries Microsoft has, the more data it has, and the more data it has, the more relevant its search results will be. Similarly, the more share Microsoft has, the more advertisers are likely to bid on keywords, and the more advertisers bid, the higher keyword prices will go.
We believe the reason Microsoft has built the infrastructure necessary to process 40%-50% of the U.S. search market is that Microsoft believes it needs this much share to become truly competitive with Google on relevance and revenue-per-search. If Microsoft is only able to capture 35% of the market long-term, we don’t think it will have enough share to become competitive with Google on revenue per search. And this will make it all the harder for the company to turn Bing into a profitable business.CONCLUSION
In short, Yusuf and the Microsoft folks have an absolutely massive challenge ahead of them. They do have the luxury of having a boatload of money ($2 billion of losses per year, at the current rate), but we doubt that even Steve Ballmer’s patience will last forever.
We believe that the dynamics of the search market are such that it favours a natural monopoly like Google, and we just don’t think Bing will be able to gain enough competitive advantage for Bing.com to grow to encompass 30% market share. And unless Bing does that, we just don’t think it will ever be a truly successful business for Microsoft.