Updated from 2:59 p.m. EDT

If you think Oracle's ( ORCL - Get Report) 18-month fight to buy PeopleSoft ( PSFT) was tough, wait until you see the struggle to make it work.

At $26.50 a share, or $10.3 billion, Oracle's victory did not come cheap; in fact, it was at the high end of the range that analysts had called accretive.

Based on PeopleSoft's own guidance for 2005, the takeover price works out to 25 times earnings. But many analysts thought the guidance was very aggressive. "Realistically, we're talking about 35 times earnings, and that's pretty steep," said Donovan Gow of American Technology Research, which does not have an investment banking business.

Winning a profit on that massive expenditure will require a well-thought-out campaign to mesh two radically different corporate cultures, retain key employees while firing thousands more, and convince customers that Oracle will support them well after PeopleSoft repeatedly warned that the database vendor would let its existing customers languish and refuse to develop new PeopleSoft products.

Still the initial reaction on Wall Street was ebullient, partly because Oracle also announced a stronger-than expected second quarter on Monday. Ironically, the brightest spot in Oracle's earnings announcement was a 57% year-over-year increase in application revenue, the usually weak area Oracle hopes to bolster by buying its rival.

Not surprisingly, PeopleSoft shares jumped $2.48, or 10.4%, to $26.43, or $2.39 a share above the stock's former 52-week high. Defying the normal acquisition pattern of a loss for the buyer, Oracle soared $1.35, or 10.2%, to $14.63.

A number of investment banking firms, including PiperJafrray and Prudential, upgraded Oracle or raised their target prices for its shares. Prudential's Brent Thill estimated that the deal will add about 8 cents a share to his 2006 stand-alone estimate, bringing his projection to 75 cents a share.

The consensus of analysts polled by Thomson First Call before the merger was 65 cents a share on sales of $11.6 billion in fiscal 2006. "Longer-term, we think Oracle will be able to post EPS growth in the low double digits as it capitalizes on new opportunities in the applications market and cross-selling of its database technology," Thill wrote in a note to clients. (Prudential does not have a current investment banking relationship with Oracle.)

Not only did Oracle pay more for the company than Wall Street had expected, it repeatedly promised not to raise the offer -- and then did so anyway. "It was a more profitable business than we had thought," CEO Larry Ellison said during a conference call Monday morning. PeopleSoft's stream of continuing maintenance fees was higher than Oracle had thought, he explained.

Sanford Bernstein analyst Charles Di Bona said, "That raises a real credibility question: How disciplined an acquirer are you?" Oracle, a company that expects to make more important acquisitions, may have trouble getting future targets to take them seriously when it says an offer is its "best and final," as Oracle execs did at a number of points during the long fight. (Di Bona's firm doesn't do investment banking, but its parent, Alliance Capital, is long PeopleSoft.)

Steve Cohen, a portfolio manager at Kellner DiLeo Cohen, doesn't buy that argument. "The circumstances were very special; you had a board that was unusually intransigent. Going through more courts fights and a complex proxy fight would have been very expensive," he said.

In any case, that's a future concern; Ellison said he doesn't expect to make another major purchase until PeopleSoft is integrated into Oracle. That remark may have been the catalyst for a quick slide in shares of BEA Systems , a struggling software company widely seen as a potential Oracle acquisition target. Shares of the company were off 9 cents, or 1%, to $8.61 on day when most software issues moved up.

Of more immediate concern to Oracle than its future M&A strategy is the massive organizational challenge posed by the PeopleSoft deal. Between them, Oracle and PeopleSoft employ over 50,000 people and have thousands of customers and scores of products.

PeopleSoft has yet to finish integrating J.D. Edwards, which it purchased immediately before Oracle began the M&A brouhaha in June 2003. JDEC, as it was called, was much smaller than PeopleSoft, and the acquisition was friendly. Oracle's task is far more difficult.

Unlike the energy sector, where mega-mergers achieve enormous economies of scale, software companies own little of value in addition to their intellectual property and customer loyalty -- and neither of those items is bolted to the floor.

It's no secret that Oracle is going to make massive personnel cuts; during her testimony in federal court last spring, co-President Safra Catz talked about cutting as many as 6,000 jobs to realize savings of about $1.18 billion. Although people may have assumed that all the layoffs would come from the PeopleSoft side of the house, Ellison said Monday that cuts will occur in both companies.

So Oracle is going to have to move very quickly to identify key employees and make sure they don't jump ship.

Customer retention may be even more important. PeopleSoft pulls in about $1 billion a year in maintenance revenue, and because it comes from existing customers, it is very high margin. Oracle has literally sworn that it would continue to support PeopleSoft's customers, but it will not develop new PeopleSoft products.

Oracle will have to sell itself to those customers. Although some defections are inevitable, Oracle will probably do well on this front given the enormous difficulty and expense a large business faces if it chooses to switch its critical business software.

One area where Oracle clearly comes out ahead is the applications business. Despite a good showing in its most recent quarter (up 57% year over year), the company has struggled to compete outside of its core database business and is much weaker than German software giant SAP (SAP:NYSE) . "In our view, Oracle moves from a weak No. 3 player to a strong No. 2 player in the application software market," said First Albany analyst Mark Murphy.

Indeed, SAP ( SAP - Get Report) is certainly the player to watch at this point. Some analysts, including ATR's Gow, figure it could be the ultimate winner if Oracle stumbles.

Even though Oracle won't be in the market for more software companies anytime soon, its victory sets an important precedent. "The conventional wisdom is that hostile takeovers don't happen in software because you don't want the employees, the intellectual capital, to walk out the door," said Tony Zingale, president and chief operating officer of Mercury Interactive which sells software that improves the performance of corporate networks and applications.

Oracle has yet to prove that it can make the merger work, of course, but the fact that it succeeded in buying the company could well change the rules in software, Zingale said.

Earnings, Forget Me Not

Overshadowed by the merger news was Oracle's strong fiscal second-quarter performance, released before the open Monday morning.

Net income grew 32 percent to $815 million, or 16 cents a share, in the quarter while total revenues were up 10% year over year to $2.76 billion year. Analysts were expecting a profit of 14 cents on sales of $2.6 billion.

Second-quarter software revenue rose 13% to $2.22 billion while services revenues grew 1% to $533 million. New software license sales were up 14% to $971 million, boosted by a 57% increase in new applications sales.

"Oracle is at its highest-ever levels of profitability," said Oracle CFO Harry L. You. "For the trailing four quarters, operating income reached a record $4.2 billion while cash flow was at $3.4 billion. During that same 12-month period we met our publicly stated goal of 40% operating margins for the very first time," he said.

It's worth noting, however, that Oracle was expecting a currency boost of 1.3%, but got 4% as the dollar weakened sharply in the quarter.

Looking forward, Oracle predicted that total revenue in the third quarter will range from $2.7 billion to $2.79 billion. EPS will be 14 cents or 15 cents a share, while license revenue will grew from 4% to 14%. The company said it is comfortable with license growth of 9%.

Analysts were expecting a 14 cent profit on sales of $2.66 billion.

In keeping with TSC's editorial policy, Snyder doesn't own or short individual stocks. He also doesn't invest in hedge funds or other private investment partnerships. He invites you to send your feedback.