There's an irony about the Microsoft-Yahoo search deal. At one time, Yahoo provided search services to Microsoft. Now Microsoft is returning the favor. Well, if anyone could call the outsourcing of Bing search to Yahoo a favor. It's not. I predict that the deal will create two losers, with Google lapping up more search share -- at least in North America.

The Microsoft-Yahoo deal is a strange one. It's the difference between two people living together and getting married. Microsoft and Yahoo will share common residence, but within somewhat separate confines, work schedules, belongings and bank accounts. This isn't even a marriage of convenience. I wouldn't even call it cohabitation. Microsoft and Yahoo are roommates who share space but different, often conflicting, priorities.

There's a screenplay in this deal, another sequel to movie "Dumb and Dumber." But which CEO, Microsoft or Yahoo, should play which role? The deal is "dumberer" for both companies, and for different reasons.

Search cannibalization. Microsoft CEO Steve Ballmer has long wanted to combine his company's search share with Yahoo's. The idea: Bring Microsoft search share closer to Google's. According to ComScore, Google's US share search share was 65 percent in June, followed by Yahoo at 19.6 percent and Microsoft at 8.4 percent. Conceptually, combined Microsoft-Yahoo would be 28 percent, which is a lot closer to 65 percent than is 8.4 percent.

Microsoft's reasoning, as first explained during the failed Yahoo takeover in 2008: Combined search share would give Microsoft enough reach to build out the customer and advertiser experience. I think everyone should have dreams, so, please, pardon my chucking black paint onto Steve Ballmer's rainbow.

Microsoft's math doesn't add up. Bing is proof of that. Microsoft search share was up 0.4 percent month on month in June, while Yahoo receded by 0.5 percent. Search share for the other top five providers, including Google, was flat month over month. By all appearances, Bing cannibalized Yahoo search during the new engine's first full month of service.

Further cannibalization is inevitable, and there is likely to be heaps of it. Matters would have been worse had Microsoft bought Yahoo and consolidated all search under a single brand. My prediction: Combined Microsoft-Yahoo share will be less than 20 percent within 12 months of the deal's closing -- and that's my being somewhat generous so that I don't get totally flamed in comments.

Crown jewel. Yahoo is giving up its core asset to Microsoft. Yahoo essentially has three crown jewels: Its popular and hugely valuable brands, search and banner advertising. Search is the core around pretty much everything revolves. Yahoo wants to give that up to a competitor? Exactly what incentive does Microsoft have, then, to help Yahoo?

Microsoft and Yahoo are competitors by most every measure other than PC and server software. Other than search, their areas of competition include: small business hosting and services, Web e-mail, instant messaging, portals and online finance, mapping, travel, health and music, among many other products and services. Yahoo's online reach, in terms of branding and time people stay online, is greater than Microsoft's.

It's inconceivable to me that Yahoo would give up something so valuable for limited gains:

• $700 million -- $500 million operating revenue and $200 million cost reduction
• $250 million increase in cashflow
• From Microsoft, 88 percent TAC, or traffic acquisition cost, at an initial rate of 88 percent of search revenue

It's like Yahoo has sold its soul to the devil under separate five- and 10-year terms. Yahoo started as a search company. Search is still core to the company's business, strategic and technological identity. Overture, which Yahoo bought in 2003, pioneered the keyword business model that Google made mighty profitable. Yahoo search share is more than double Microsoft's for good reasons. Search is Yahoo's soul.

The crown jewel Yahoo gives up today is another it loses tomorrow. Future search technology development will come from Microsoft, not Yahoo. What is smart about Yahoo giving up search technology development and essentially abandoning its research and development investment in the Panama ad system?

Something else: Bing search share gains, as widely reported by many news sites, are overrated. Those weekly search gains cited after Bing's launch aren't yet sustainable, as shown by more granular search data released by ComScore. There were 1.02 billion searches at Bing in June. Windows Live/MSN racked up 1.19 billion searches in May, 1.25 billion in April and 1.19 billion in March, according to ComScore. From that perspective, Bing search declined month on month in June compared to Windows Live/MSN in May. So let me see if I understand. Yahoo wants to swap out something that works for something unproven? It's another reason to wonder about Yahoo losing share once Microsoft controls search.

Missing mobile. Microsoft has been chasing Google for way too long. I would dare call Steve Ballmer obsessed with catching Google. Microsoft is distracted from more important matters, such as developing a solid mobile strategy, which includes search. I keep wondering how much of this Google obsession is about pride. It's like Microsoft executives are trying to recapture 2001-2002, when Microsoft led Google in search.

Two events catapulted Microsoft to US search leadership, starting in late 2001: Internet Explorer pushed "404" page not find errors to MSN Search; Windows XP launched. Microsoft not Google was the search leader. But by the end of 2002, Microsoft had fallen to third rank behind Google and Yahoo. MSN Search kept steady, while Google and Yahoo gained users. According to what was then ComScore MediaMetrix: Between March 2002 and May 2003, MSN Search grew by about 1 percent, as measured in number of visitors, compared to about 25 percent growth for Google.

Microsoft might regain some glory days, if the company spent more time on mobile, which is the future of search. But Google has won the desktop and is rapidly gaining on mobile devices. Cell phones are captive devices, meaning most people carry them -- and increasingly use them for more than telephony. In June, ComScore reported on March-to-March mobile search usage: Up 51 percent, in the United States. Something else: ComScore said that mobile Web browsers are the "leading access method for seeking local information."

Meanwhile, mobile advertising is in its infancy, but growing. Google doesn't have to dominate mobile the way it does the desktop. Last week, the Kelsey Group released its forecast for mobile search and advertising revenue in Europe. Between 2008 and 2013, revenues are predicted to increase from 39 million euros to 2.3 billion euros (or US $55.49 million to $3.7 billion, at today's exchange rate). The European mobile market is larger than the United States -- 499 million handsets versus 266 million, respectively, at the end of 2008, according to Kelsey Group. The analyst firm sees smartphones as being a major driver of mobile advertising. The number of smartphones in Europe is predicted to increase from 32 million last year to 149 million by 2013.

Microsoft needs a mobile operating system and browser Manhattan Project a lot more than it needs a Yahoo search deal. If Steve Ballmer must obsess about Google, he should focus on where the rival isn't yet entrenched and where also is the future of computing.

The antitrust quandary. The Microsoft-Yahoo deal is subject to regulatory approval. Those trustbusters must have their say. Some lawyer surely will argue that the deal could lead the market to consolidate around two search engines: Bing and Google. The other two major search leaders, AOL and Ask, had 7 percent combined search share in June, according to ComScore. Their share will recede, someone will tell trustbusters.

Now here's a strange thought: What if regulators reject the deal not because Microsoft-Yahoo would be too big but because search cannibalization would make Google too big? That's not typically how antitrust law applies, but there is an argument for the deal making Google a bigger monopoly. Strange as it might sound, Google lawyers could use this argument against the deal. It is not typically how companies argue antitrust -- that a deal should be prohibited because a dominant third-party would gain more market share. Nevertheless, the logic is sensible.

Google already is a bigger monopoly. AOL's search share was 3.1 percent in June. But, technically speaking, that's really untabulated Google search share; Google still provides paid and unpaid search to AOL. So, Google's real US search share nips 70 percent, and paid search is nearly 80 percent. Do trustbusters really want to risk making it more?

Now, I suppose that Microsoft could seek to steal away AOL, which deal with Google is up for renewal. Then Microsoft would get that untabulated search result. Such circumstance would be good for Bing, I predict, even pushing total search share over my predicted less-than 20 percent.

I'll end with that. I could write a short book on why this deal is bad for Microsoft and worse for Yahoo. Well, damn, I gave away which is Dumb or Dumber.

By the way, as is typical of this kind of announcement, the analysis is my own. I reviewed the announcement material and the BetaNews "need-to-know" bullet-points story. That way, no one else's thinking influences my opinion.