IBM and the Art of Acquisitions
“If you do something 60 times you are going to develop some best practices,” Scott Hebner says.
This could be said of many things. The thing Hebner is talking about is acquisitions—software company acquisitions by his employer, IBM, to be exact. Hebner is vice president of marketing and strategy for IBM Rational Software, which makes tools for software developers and is itself the legacy of an IBM acquisition, five years ago this February. So he should know what he’s talking about when he says that IBM knows what it’s doing when it buys a software company. After all, as he observes, “If there’s one thing IBM is good at, it’s putting processes in place and operationalizing them.”
So, 60 times in the last dozen years—more like 66 times, actually, counting the newest case just last week (read on)—IBM’s software division has chosen to expand its business by acquiring other companies that were offering products IBM wanted to offer, or that had capabilities IBM might have built internally but found it easier to buy. And it’s getting pretty good at it. The word is that some of the early acquisitions—Lotus (1995) and Tivoli (1996) come up in conversations—were a bit rocky. But now IBM has an entire team of people who do nothing but manage the assimilation of newly purchased companies.
It’s a balancing act, but one IBM seems to have mastered, says Mike Weider, founder and CTO of Watchfire, a security company that joined IBM last year. “If you integrate too quickly you smother the fire that made [the acquired company] special,” Weider says. “But if you don’t integrate fast enough then you don’t leverage the synergies. I think IBM has figured out a nice balance between those two things.”
Some of those acquired companies have now been around long enough inside IBM that they’re making acquisitions of their own—many of them right here in Massachusetts. Waltham, MA-based Watchfire, for example, became part of Rational. (In many cases, including with Rational and Watchfire, IBM keeps alive the brands of its acquired companies—usually when the brand signifies something to the market that IBM’s brand by itself couldn’t convey.) And in the deal announced last week, the newest member of the IBM software family—FilesX, a Windows data protection and recovery company with headquarters in Newton, MA, and Haifa, Israel—will become part of the Tivoli organization.
None of this is to say that IBM has acquisitions down to a science, because every acquired company is different. The impression I get from speaking with people like Hebner and Weider over the past couple of months is that the company sees acquisitions as something closer to an art—but one that gets easier, as most artistic pursuits do, as you develop a mastery of technique.
“There are very few companies out there that are able to acquire companies and do a good job at integrating them,” says Mohamad Ali, an IBM vice president for business development and strategy who is Big Blue’s transition manager for the acquisition of Ottawa, Ontario-based business intelligence company Cognos (which has a major office in Burlington, MA). “IBM is one of a handful. I don’t say that just because I love it here. Statistically, if you look at the acquisition process, about 30 percent of acquisitions actually hit the target, in terms of the synergies that create the premium that make you want to acquire somebody, which is really quite dismal. At IBM, our percentage is much higher.”
So what are the keys to good technique, when it comes to buying a company and everything that goes with it? Obviously, as EMC executives detailed to Bob recently, there needs to be a good technological fit between acquirer and acquiree. According to Hebner, IBM Software doesn’t buy a company unless it has already determined that its customers have a need; that an outside company’s software fills that need; and that it makes more sense to buy that company than to try to build an equivalent system from scratch, or to cobble something together from IBM’s existing products.
Watchfire is a good recent example. “Rational has products that span the software development lifecycle, from architecture and modeling to coding tools, but did not have many capabilities in the area of security,” says Weider, who founded the company to build software that automatically evaluates websites and Web-based applications for security holes. “We are really filling an important need in there. To really build security into a company’s software, you need to get engaged with some of the development tools they’re using.”
But that’s the obvious part. Companies generally don’t go around spending hundreds of millions of dollars on acquisitions—or billions, in the cases of Lotus, Rational, and Cognos—unless the company being acquired has complementary products that fill a real need in the market. What’s just as important to IBM is that there be enough overlap with its own products that it will be easy to plug the new company’s widgets into Big Blue’s global sales operation. That way, the acquired company gets to grow even faster inside IBM than it would have if it had stayed independent.
“Watchfire had their best quarter by far in the fourth quarter” of 2007, says Hebner. “They’re based in North America, but they landed new deals in India and other parts of the world because we were able to get them into some of the bigger accounts we have access to. Taking a company with a strong value proposition and getting them integrated with our worldwide sales organization—that alone is fuel for growth.” Similarly, Rational’s business has grown 40 percent since it joined IBM.
But synergies between products don’t mean much if the people who actually make them aren’t happy. So IBM says it puts a lot of thought into the human side of the acquisition equation. “You are acquiring the people—that is the asset in most of these cases,” says Hebner. “The code base, without the people who understand what it does, is not very useful. So we’ve put a much bigger emphasis making sure that the people are happy. We do a lot to make sure that we celebrate the culture of the company coming in. We actually try to assimilate things from them—and IBM has become a much more flexible place because of all these new people.”
As an example, Hebner points to Rational (which IBMers often seem to hold up as the company’s model acquisition). “Rational was very entrepreneurial and practitioner-oriented—it sold directly to users,” explains Hebner. “If it was a tester product, they sold it directly to testers. To do that, you’d better be technical—and [Rational founder] Grady Booch is the perfect example, very much a thought leader, very deep in the skills and the business processes that customers go through to deploy software.” IBM has tried to absorb that mindset and approach, says Hebner. “We have a much better ability now to talk to developers in an intelligent and relevant way than before Rational came on board.”
IBM’s respect for the talent it’s hiring is not lost on the talent themselves. IBM “has a reputation for doing a great job with acquisitions—which is not necessarily common in our industry,” says Weider. “When you think about why acquisitions fail, it’s typically one of two things—either the strategy was off or the execution was off. In our case, the strategy is bang on.” On the execution side—well, so far, eight months into its assimilation, Watchfire has retained 99 percent of its original employees. “That’s pretty much the largest proof point that things have gone well,” Weider says.
If there’s a problem with becoming part of IBM, says Weider, it’s that the 270,000-employee behemoth is almost too welcoming: Everybody around Big Blue wants to know all at once how they can help the new kid, and how the new kid can help them. “The challenge is being overwhelmed by all the different people who want to leverage your people and technologies,” says Weider. “I spend a lot of time being the spokesperson for our division within IBM, and if I responded to every single request, I would be doing nothing else. You have to know what’s critical and how to prioritize things.” Luckily, that’s a process question—and IBM seems to have gotten pretty good at those.