Steve Ballmer (Microsoft photo)

From a financial perspective, Microsoft's new search partnership with Yahoo doesn't even come close to the $45 billion acquisition bid made by the Redmond company last year. Microsoft expects to spend somewhere between "a couple hundred" and "several hundred" million dollars to integrate the technologies over the next couple years, Microsoft CEO Steve Ballmer said this morning.

In return for running the underlying search and advertising technology for Yahoo sites, Microsoft will need to pay 88 percent of the revenue back to Yahoo in what's known as "traffic acquisition costs." That means Microsoft will keep 12 percent of the revenue. Estimating Yahoo's search advertising revenue at roughly $2 billion a year, that translates into $240 million for Microsoft.

[Update: In a note to clients, Goldman Sachs analyst Sarah Friar puts Yahoo's search advertising revenue more in the range of $3 billion a year, or $360 million a year for Microsoft, a higher number but still in the range of the costs Ballmer expects.]

In other words, in the short run, this could essentially be a wash for the Redmond company, looking at the costs and benefits. But in the long run, Ballmer is hoping the partnership will pay off significantly in other ways. Here's how.

First, the deal would channel a much larger volume of queries through Microsoft's Bing search engine. That promises to give Microsoft much better data to work with as it tries make the paid search advertisements more relevant to the terms people are searching for, said Matt Rosoff, an analyst at research firm Directions on Microsoft.

For example, Microsoft would be able to know generally which search ads Yahoo users click on, within privacy guidelines cited by the companies, to improve its approach.

At the same time, Microsoft's underlying AdCenter technology would power the combined Yahoo-Microsoft search advertising system -- benefiting from a much larger pool of advertisers. With more advertisers bidding on keywords, that will result in generally higher keyword prices, within limits.

But in a broader sense, "this is as much about making sure that Google doesn’t grow any more, as anything else," Rosoff said. "I think that’s what they’ve been going after for the last two years as they've been trying to do this deal."

Here's what Ballmer said this morning when analysts asked about the financial implications for Microsoft.

We paid a high (traffic acquisition cost) rate, there's no question. In the short run there's a little bit of transition cost, a little bit -- there will be several hundred, a couple hundreds of millions probably over about the first two years of implementation of transition costs that we're betting into the future. T

he upside for us really comes as we're able to improve the relevance of our search product by having a bigger -- because ads are part of being relevant and we have a chance to do a better job of relevance. And then as we can improve the monetization on both the Yahoo site and the Microsoft site, there's good opportunity for us for upside in that.

So we clearly see an upside, but our upside comes as execution really builds.

Carol gets some upside in terms of cost in the short run, and then gets this upside. For us, the investment in the near term will be a few hundred million over the first couple of years, because there's a lot of work, a lot of work that's involved in this transition to our platform and we're stepping up to that, but there's a lot of expense associated with it.

We have some extra capacity we'll use in terms of the physical data centers. There will be some initial (capital expenditure) as we get into the thing, nothing that affects ... the FY10 period for us. You gotta remember one of the reasons our capex is down year-over-year is we're not building as many buildings in Redmond. Independent of data centers that's not affected by anything we're talking about here.

So from an (operating expense) perspective, we have a plan in which over time some Yahoo engineers may move to Microsoft but that's kind of in part of our budgeted thinking, we'll just have better talent, if you will, to help us, but we definitely will have to invest proportionate to volume, and the ongoing increase in the number of Web documents that need to be indexed, etc. Some of that will certainly be incremental to what's in our current plans.

I don't think we will knock you off your feet with any changes we may make in our guidance, and it won't be for FY10.

Previously: Microsoft and Yahoo strike deal: How it will work, what it means