How Semantic Search Impacts Ad Targeting
Paid-search marketers should see far better ad targeting once Google and Bing build out semantic technology in their respective search engines. Colin Jeavons, CEO of Vertical Search Works, said targeting falters today because of a phenomenon he calls the Big Band Theory.
"As more pages of content get created daily, it's like the expanding universe," Jeavons said. "The keyword search model becomes more difficult, especially when trying to serve up accurate results on queries. With semantic search you're not striking the query off one keyword, you're striking it off multiple -- to the tune of millions in some cases, which means you can render a better match to a query with semantic search."
Google's algorithms run more on probability rather than working like a brain, which takes the content of multiple words to find meaning, Jeavons explains. From a technology perspective, there's no question that "Google is very weak in many areas," Jeavons said, refereeing Google Fellow Amit Singhal's Google+ post last week explaining that the engine would increase a focus on semantic search -- a technology Jeavons has been analyzing for years. The technology aims to improve search accuracy by understanding the intent and contextual meaning of searches on engines.
One of the major challenges for Google, Jeavons said, is that the engine serves many text links that aren't really accurate. Searchers don't typically click on one blue link, but rather five -- so in some ways, he believes, Google's inefficient technology generates more revenue for them. The question then becomes: will a more efficient search and ad platform create less revenue for the company?
But Google really doesn't have a choice. It must change. Jeavons believes that today, Bing integrates more semantic technology than Google.
It wasn't always that way. During the first revolution of search in 2000 and 2001, Google produced a far better product than any others, according to Jeavons. Yahoo, its main competitor in 2002, focused on becoming a portal and distribution model, losing its way in search so its nearest competitor didn't have to improve the quality of its engine to dominate search worldwide -- not until Microsoft released Bing. Jeavons believes the release of Bing woke up Google. Search one phrase on both Bing and Google and get two different query results.
Even Singhal admits it. He wrote in his Google+ post that Google's ability to understand questions "is pretty darn limited." Jeavons agrees. He said Google didn't achieve "world domination" for its technology, but rather gained it because of the lack of competition. The actual challenge is not technology and quality, but distribution and brand.
"Is Coca-Cola a great brand, or simply something consumers want in developing countries and the manufacturer figured out a model to distribute the brand better?" he said.