INTRODUCTION

Search engines are one of the most widely used Internet applications. According to Purcell et al. (2012) ‘91% of online adults use search engines to find information on the web...

Advertising is, in general, a low-yield business. Ad rates are typically measured in cost-per-thousand (CPM) impressions. A typical CPM for a prime-time TV ad would be US$10, which translates to 1 cent per impression per user. Online display ads might sell for US$1 or US$2 CPM, which is an order of magnitude less than TV ads. Search engine ads, by contrast, could easily have CPMs in the range of US$100. Why? Because the ads are much more relevant to users’ interests since they are tied to user queries. Even though the yield of search engine ads is high by comparison to other forms of advertising, it is low in absolute terms. A good ad click-through rate might be 3 percent and a typical conversion (purchase) rate might also be around 3 percent. This implies that fewer than one out of a 1000 people who see the ad are actually induced to buy the product being advertised.

The fixed costs of entering and running a search engine were at one time substantial since the entrant was required to build or lease a data center. Nowadays, those fixed costs have become variable costs due to the availability of cloud computing services such as Amazon Web Services. This has enabled the proliferation of special purpose search engines that focus on particular types of searches such as travel and shopping.

However, it must be acknowledged that general purpose search engines tend to build and operate their own data centers and so incur significant fixed costs. Companies that already operate such data centers are in a good position to offer general purpose search services. Examples would include IBM (Watson is, in fact, to large degree a search engine), Facebook and Apple. The latter two companies already have a substantial user base for services that they may choose to offer in the future.

On the demand side, user switching costs for search engine users are very low: the competition is just a click away. Fallows (2005) indicates that 56 percent of search engine users use more than one search engine. Hence, we can expect to see robust competition for users among incumbent search engines.

Not only are users not exclusively tied to a single search engine, neither are advertisers. Typically advertisers will ‘multi-home’ and advertise wherever there are enough potential customers to warrant the relatively small costs to port ad campaigns. These characteristics

385

of the search engine business - low marginal costs, low switching costs, and an advertiser- supported business model - means that the likely market structure for general purpose search engines will be one with a few large competitors in a given country or language group.

On the other hand, there could be a multitude of special purpose search engines. There are special purpose search engines for travel, shopping, local services, academic arti­cles, and so on. The equilibrium industry structure might be similar to that of national newspapers or news magazines: a few large providers, supported mainly by advertising with ongoing competition for new readers. There are no significant network effects or demand-side economies of scale that would drive the market to a single supplier in either the search engine industry or the magazine industry.

Another analogy would be to think of department stores as general purpose shopping venues with shoe stores, men’s stores, cosmetic stores, furniture stores, and so on, offering special purpose shopping services.

I will argue later that the most important economic factor determining search engine success is learning-by-doing (Arrow, 1962). Because of the low user switching costs, search engines have to continually invest in improving both their search and their mon­etization. Though this could be said to be true of virtually any product, continuous improvement is particularly important in online products since pace of experimentation and implementation is particularly rapid. Though there are dozens of search engines available, the big three in the USA are Google, Yahoo! and MSN. I will mostly discuss Google, since I am most familiar with its practices, but the other search engines tend to use similar business models.

18.2