ChangeMedia

Those in the know love to explore Google Labs, the playground of future Google technologies not quite ready for the world. Some of the Google Labs projects will make it to the big-time, some won't. A recent addition to Google Labs site is Google Suggests. Google Suggests has been much noted for its cool technology. By my readings, not much has been noted about why it is either good for me as the consumer of search tools or for Google as the profit making money collecting soon-to-be-vast-web-empire. I argue below that Google Suggests is perhaps more about optimizing the value of our searches from the Google perspective than it is for optimizing you and I getting the results we are looking for.

Firstly, Google has developed a great brand - we almost all believe at this point that the value relationship between Google and ourselves is more tipped in our favor than it is with some other service providers. We're not having late night lovers quarels with Google complaining that we're putting more into the relationship than they are. Right? I can only asssume that Google really, really, really wants to protect this brand. One question I will pose later is whether Google suggests will have a positive or negative affect on the brand that is Google.

The value of Google Suggests is weighted in favor of Google, not me and not you. How can this be? Isn't it really great that I will require fewer page loads to get to the results I want? Isn't it great that I'll be sure to search for the terms that will return the largest number of results? Isn't it great that I'll never misspell another search term again? I would argue that its more great for Google than it is for you and I and it may even be Net bad for you and I.

Before I get going, you must know how Google makes money from your searches, if you don't already (skip to next paragraph if you do). Advertisers purchase ad placement next to results for specific search terms. When you or I search for one of these terms our results include (I use "include" since recent articles and studies suggest that most (!) people can't tell the difference between the ad and the search results) an ad. If you or I click on that ad, Google gets paid. If we don't, Google gets the big donut, diddly, zilch, nada. For instance, if I go to Google and search for "blog", one of the ads that comes back is for "Type Ad". Type Ad has contracted with google to place an ad next to search results for search terms that include the world "blog". Type Ad then pays Google if I click on the placed ad.

On to the more meaty stuff.

I classify every search on Google in one of three ways: 1.) direct hits; 2.) direct miss and 3.) missfired. The terms are very much from the perspective of Google's balance sheet, not your or mine's search experience. Also...please pardon the gun metaphor - its disgustingly awful. Some definitions:

Direct Hit is a search term for which google has sold an Ad. I'm assuming here that there are also search results returned. At the very least, this search has the potential to generate revenue for Google. It still might not...but its a great start.
Direct Miss is a search term that returns results but for which no advertisements have been sold. A direct miss is valuable to the searcher, but not valuable to Google.
Misfired is a search where the user just plain old screwed up, didn't get search results at all because they typo'd, spelled wrong, forget to enter anything at all etc. Importantly, these typically (I'm assuming) result in the searcher trying again and ultimately landing on one of the other two above.

So...in this context, I argue that "Google Suggests" is a search value optimizer, again from the perspective of Google. It encourages - or will ultimately be used to encourage - a set of searches that on average result in more Direct Hits and fewer Direct Misses and Missfires. I'm all for eliminating Missfires (and as such really enjoy the "Did you mean?" feature in Google which does not avoid the missfire, but helps me quickly resolve it). But...unlike the Misfires, Direct Misses are only misses for Google - for me they are often generate results from Google's full text index that are exactly what I want.

If I'm Google and I can get users to more closely congregate their searches so that that similar concepts are phrased identically then I've done some good. So...if searches that start with individuals thinking to themselves "San Francisco Lodging" or "San Francisco Motels" or "San Francisco Hotels" all end up being entered as "San Francisco Hotels" then Google has created more search concentration around this keyword and they'll start seeing higher or more bids for those keywords rather than seeing keyword purchases fragmented around various alternatives for places to stay in San Francisco. I would expect this to happen since as I'm typing "San Francisco" the fourth or fifth suggestion down is "Hotels".

Also, if I start typing "San Francisco" and that was going to be the end of my search phrase then I might click on an ad and shazam, I'm off to the San Francisco Hilton site (one of the ads associated with "San Francisco") . You can bet that Google would have preferred that I search for "San Francisco Hotels" instead of plain ole' "San Francisco" since the cost of the more specific search phrase "san francisco hotels" is very likely higher than the more generic "san francisco" and thus creates more Google $. (take my word on the pricing of keywords in Google - "San Francisco Hotel" is much more expensive than "San Francisco"). If I'm looking for a hotel, but am the average searcher aren't I more likely to append that "hotel" option to the end of the phrase if its suggested, than if it is not?". In that scenario the search has been OK for me and VERY good for Google.

A far more interesting scenario is one where I intend to search for a search term for which no one has placed an ad. If I do indeed get back a result and it is one that I want, google has missed the opportunity to extract value from my search. I speculate the aggregation of searches around specific terms noted above is far more powerful in terms of revenue benefit to Google than revenue lost to unsold search terms...but I'm not sure (if you know, drop me a line). A question to ponder about the "average searcher" is whether he or she would second guess a search term when google had nothing to suggest and "try again" before executing the search? I probably wouldn't....but....I bet lots and lots of people working on the thought that the absence of a matched suggestion means that no or few or invaluable search results will come back.

Also...ponder these thoughts about searching generally and how they match up with Google Suggests "technology":

Does the suggestion including the number of results help me? Isn't the goal of most of my search to return the minimum number of results that encompass the stuff I'm searching for? The presence of the number kinda suggests the opposite, don't you think? Usually if I search for something that returns a bazilllllllion results I'm only going to be interested in the first few based on page rank or I'm going to search again.
Isn't effective searching from the searcher's perspective best done as an iterative process of search, view results, tweak terms, search view results etc. etc. etc.? Is there something wrong with this process? There is some sort of learning that goes in while engaged in this process that is lost when I'm told that when I think about San Francisco that what I'm really "supposed" to be thinking about is San Francisc XXXXX where XXXX is determined either by Google or by the rest of Google's users and NOT the nature of the internet world exposed through Google's Index. If it ain't broke....

Isn't this a step backwards folks? Isn't this a glorified site directory ala Yahoo 1994 that just happens to have an interface that starts with me typing?

I leave you with a few questions:

Is the decrease in Missfires and increase in Direct Hits with some cost to the consumer in the subtle effectiveness of freely entered text searches worth it?
Will searchers ever perceive the narrowing of both searches and search results around the likely more limited set of suggestions that they willing elect from the pulldown. If so, will narrowing create an opportunity for a more generalized free-text search?
Will I capture 1/26th of world's search revenue if I create a website directory called "Search terms that start with the letter A"? I'll leave room for 26 competitors. If you feel like you can improve on my idea, you may want to consider creating a directory site called "search terms that start with the letters AR". In fact...you may want to create a directory of such directories then we can as a distributed system mimic the functionality of Google suggests perfectly. :)
Will google ever release to the public "Google Suggests"? Is there a way to tailor Google Suggests so that it is truly valuable to both parties?
If you are skeptical of my analysis, do you think I"m 1% right? Would google do it for 1% more revenue? For 1% few misfired (overhead) or Direct Misses (overhead without potential)? For 5%? More importantly, is Google now a company that would sacrifice customer goodwill for that 1%? 5%?

Its an interesting feature - and very well implemented. Is it of value to me? Is it of value to Google?