In strategy class you learn that when a company in a market is making “too much money” (i.e., above normal, risk-adjusted returns) that new entrants will be attracted to the market, unless there are factors that prevent them from so doing (e.g., a regulated monopoly).
One question I’ve often wondered is: why hasn’t Google attracted more competition? To answer it, let’s first remind ourselves of how Google got to where they are:
- Competitors took their eye off the ball. First-generation Internet search engines took their core market for granted. They decided that “search was a feature,” not a market, and that the real market was portals. So they focused their energies on licensing content (from news to horoscopes) and outsourced the search box to others. Oops.
- They had a better algorithm. PageRank, inspired by the academic notion of importance based on citation, was simply better than word-frequency algorithms in determining relevancy for Internet search (and not, by the way, for enterprise search).
- They had a better business model. Ironically, Larry and Sergey’s hatred of advertising lead them to the best Interent advertising business model. On the user side, they borrowed an idea from GoTo (paid search results) and implemented text-only ads, clearly separated from normal search results. On the advertiser side, they spurned the notion of a direct salesforce, instead using an e-commerce auction to sell ads.
I mention these things both because they are easily forgotten in this Google-worship era, and because I’ve started to see signs that Google is drifting away from them. I’ve heard that in the future ads may no longer be text-only. Plus, the placement of ads on top of search results is a step away from the original idea. (I can almost smell the MBAs at work.)
Clearly Google meets the Strategy 101 test of making too much money, so you have to wonder why they aren’t facing more market entrants. Could it be that people have decided that Internet search is fated to be an oligopoly dominated by GYM (Google, Yahoo, Microsoft)? With Google jumping into everything from maps to classifieds, could it be that most would-be competitors are playing defense, not offense? Could it be the barriers to entry are too high?
Despite these questions, some vendors have entered Internet search in the past year or so. Let’s examine them and their strategies:
Who is trying?
- Answers and Brainboost were both launched on the idea of returning answers, instead of links to them, in response to queries. While this is a good idea, it is very hard to do against content without rich markup. MarkLogic provides this ability, by the way, because we do not merely return links to documents, but can dynamically extract and re-assemble pieces (XML elements) of them. Without good markup it’s very hard to actually provide answers, and thus neither Answers nor Brainboost are terribly effective. Google now provide a limited answers capability directly from the search box (type “how many people live in CA”) and, with a different twist, both Yahoo and Google provide a marketplace for real people to answer other peoples’ questions.
- Clusty had a nice PR launch, but they are one-trick pony. They are a meta-search engine whose one trick is dynamic clustering. While useful, it doesn’t actually eliminate the need for taxonomies and they are falling prey to the old trap of “use your Internet search engine as a demo” for something else. (Think DEC and AltaVista as the demo for the ill-fated Alpha chip.) These guys are a feature, not a company.
- Amazon is whacking at the problem in two ways. One is with their oddly-named A9 Internet meta-search engine, the other is with their Alexa subsidiary. A9 lacks a single killer feature and adds several smaller ones. A9 sources from Google, Amazon’s book search, the Internet movie database, Guru net, and others. It has nice history and organization/annotation features. By far the sexiest feature is the A9 maps blockview. Despite some nice features, A9 doesn’t seem to be getting much traction. Alexa, however, has launched a very creative, indirect attack on Internet search by opening their crawl to the public through web services. For a modest fee, virtually anyone can leverage Alexa’s infrastructure to build a new type of search engine. This radically drops entry barriers and could change the industry.
- Kartoo is out there with a visual meta-search engine. I’d rather try to buy an Xbox on Christmas Eve than try to make money peddling hyperbolic trees.
- Vertical meta-search seems to be the preferred attack route. See Trulia in real estate, Kayak in airplane tickets, or Zoominfo for people. These folks try to out-Google Google in one specific domain (think “long tail” theory) — and ironically often scrape Google results as a data source and sell Google ads for their revenue.
But the inspiration for today’s very long posting is a new company: Gravee. They’re another meta-search engine, but this time with two twists.
- They’re changing the business model. They argue that Google is a parasite on the system, charging for search results on other peoples’ content. They say they will share their revenue stream with the content owners.
- They’re changing the algorithm. They claim to have a proprietary relevance algorithm that works against the aggregated results (typically, the Achilles heel of meta-search) and that they also factor human input into the equation. Details are light on the website as to how it all works.
We’ll see how well it goes for Gravee. I’m not arguing they’re going to change the world, but it’s nice to see someone start challenging Google on the two core items that launched them to fame: the business model and the algorithm.