In the following post on his blog, Henry blodget and I exchanged some ideas on the search battle and why is in a really precarious situation.

In relaunching, Barry Diller and the IAC crew are trying to suggest that the quality of their search is better than other outlets. Part of the battle for search revenue will be convincing users that your search is best, but relying on this and this alone will be a failed strategy. The search battle is really a battle for the desktop. Let me explain why this is so.

People use search engines in two separate and distinct ways. The first is to search out unknown sites as part of research, information gathering, or to just find interesting sites about a particular topic. The second use of ‘search’ is more of a directory service. That is the case when the user knows exactly what site(s) she is looking for, but uses a search engine to provide a directory (or navigation) function. If I’m going to book a flight I could type “” into the address bar or I could type “travel” into a search box. Due to the distribution of toolbars in browsers and search bars on other sites, their role as ‘homepage’, these search boxes are everywhere and they’ve made it quicker to navigate the web to known sites using their services than they have through bookmarks or a browser URL bar(plus they’ll fix your tyops for you too!).

I don’t think that any of this is news to anyone. What’s interesting about this as it relates to the business side (as opposed to the user experience) is that 70-80% search keyword revenues comes from about 20% of the queried words and the 20% of those queries that drive 70-80% of the revenue from search engines are from users who already know the sites they want to go to. In other words, most of the real economic value in search, lies in its role as an efficient navigator to known, highly trafficked sites. This is the stuff that matters. A relevant example to prove this point comes from FTD. Say its Valentine’s Day and you want to send flowers to your wife. Most users that are ready to buy are going to search one of about 10 keywords: (flowers, flower, roses, FTD, 1800flowers, proflowers, 1-800flowers, teleflora, etc). By and large, users either know they want to buy from one of a handful of sites or they have a few sites in mind and search for a generic term like “flower” or as if they were looking something up in the Yellow Pages. Either way, while the flower websites will purchase 1000s of keywords 80% of the transaction volume and the search engine’s revenue comes from a few terms. These are the terms whose rates get bid high in the search auctions because they are both the leading drivers of volume and are also the strongest indicators of purchase intent. This basic concept repeats itself over and over again across all of the major categories in the search space.

What’s relevant here is that users don’t need a robust search engine to return FTD, 1800flowers, Teleflora and Proflowers to a user who queries “flowers”. A good search engine simply needs to prevent spammers from taking over the top spots (which BTW can be accomplished relatively easily and at a significant profit by simply putting the paid links at the top of the engine and letting the marketplace determine relevance), and they need to get the user from their current location to the results as quickly as possible.

Ask anyone in the know, and they’ll tell you that the quality of the major search providers is quickly converging. I still believe that Google is superior to the others when I’m doing an obscure query (based on the size and quality of their index), but for the navigation purposes described above, all of them are basically the same. The key in winning share of revenue in search therefore, is to make the navigation process above as quick and efficient for the end users as possible and that means keeping your search box in front of users as often as possible and convincing those users that your search engine is just as robust as the next guys.

Microsoft’s edge in the battle will be its launch of Vista and IE7 later this year where it will undoubtedly embed its own search boxes throughout the UI. Yahoo is relying upon its IM, Email, and MyYahoo users to maintain its presence on the desktop, and Google’s strategy seems to be clear as it launches Google Pack and upgraded toolbars, announces OEM deals, new software applications, etc. Microsoft is going to frame search as central to the UI anytime you are in front of a keyboard. Google is going to try and convince users that they *are* search and clearly they intend to remove Microsoft’s advantages by reframing exactly what constitutes the *desktop*. You could call it a battle of mind share, but the winners are going to be the companies that can keep a persistent presense in the UI while convincing the public that their search is best.

None of the current search buzzwords (social search, personalized search, clustering), etc matter. Those are all things that address edges of the market. For the core piece of search revenue it comes down to controlling the UI and until a robust speech recognition solution enters the market the game will be won by the company that competes most effectively in controlling the new desktop.