Marissa Mayer may try to beat Google at search - just like these companies failed to do

Ahead of today's Yahoo earnings report, Re/Code's Kara Swisher reported that Marissa Mayer is considering using the newly negotiated terms with Microsoft to rebuild Yahoo's search engine.

In 2009, Yahoo outsourced search results and search ads to Microsoft. Last week, the two companies signed a new deal where Yahoo would only be dependent on Microsoft for 51% of search and search ads.

Mayer could outsource the remaining 49% to Google - where she helped build the original search juggernaut as an early employee - and perhaps make more money.

But Swisher says that Mayer is thinking of taking on Google more directly. How? "We'll have a better algorithm," Mayer reportedly told one person.

Anything is possible!

But a lot of companies have tried to take on Google in search. Nobody has come close to dethroning the king. Here's a recap:

Microsoft first launched a beta of its own homegrown search engine in 2005 (prior to that, search results were provided by Inktomi and others). Not 2009 - that's when the company rebranded its search efforts, which had previously been called "MSN Search" then "Windows Live Search" - as Bing.

In other words, Microsoft has been working on search now for more than 10 years. Microsoft has the benefit of billions of dollars a year in profits and free cash flow, has a dedicated research arm packed with Ph.Ds, and has built a massive data center infrastructure capable of analyzing billions of web links and search queries and returning a result in milliseconds.

The end result of all this? Bing has about 30% market share in the U.S. if you include Yahoo's sites, 20% if you only include Microsoft-branded sites, according to Comscore. Worldwide, the figure is more like 8% for Bing+Yahoo, and 4% for Bing alone, as per Statista.

Between the middle of 2005 and the middle of 2013, when Microsoft broke out its online division's financials, the company lost $12.36 billion on Bing and other consumer online services.

DuckDuckGo is a search engine that differentiates itself from Google by not tracking users in any way - it compiles results from sources other than Google, including Bing, Wikipedia, and Wolfram Alpha, and shows all users the same search results. It does not store users' IP addresses and does not use cookies except when necessary.

It was founded in 2006. On Monday, DuckDuckGo passed 9 million search queries in one day for the first time. By way of comparison, Google serves that many queries in about 3 minutes and 45 seconds.

Blekko was a search engine that promised to deliver better results than Google by excluding low-quality content from so-called "content farms" and other sources and relying heavily on human-curated results. It raised more than $60 million from hotshot Silicon Valley investors like Marc Andreessen and Ron Conway, but never took off.

In March, IBM bought some assets of Blekko and folded them into its Watson technology. The search engine is no longer online.

Wolfram Alpha, which launched in 2009, is similar to Blekk0 in that it pulls information from a set of curated sources like Wikipedia, but can also compute answers to mathematical and other types of questions directly. Instead of guiding people to those web pages for the answer, it simply delivers the answer right on the site. It's used by Bing and Apple's Siri, and has branched out into other areas such as delivering analytics for Facebook. But it's still essentially a niche product.

There are also dozens of ways to conduct specific types of searches, from huge players like Amazon, which is a formidable rival to Google on product searches, to niche search engines like CC Search, which lets you search for images that have been granted to the public domain via a Creative Commons license.

Perhaps that's what Mayer is talking about. A generalized search engine is a lot harder to build than one that works for specific functions - like, say, searching Yahoo Finance when a user poses a query that seems to be related to financial information, or Yahoo Sports when users are asking for a sports score.

Mayer also noted on Tuesday's post-earnings conference call that the company is particularly interested in mobile search that sifts through personal information such as email rather than Web pages. That type of search product has a very different profile in terms of investment than traditional Web search engines, she noted.

But if Mayer is thinking of taking on Google head on in generalized algorithmic search, she'll be doing something that nobody else has succeeded at.