How Tivoli Acquired IBM. On Pulse 2012 and Cultural Change for Smarter Computing.

I couldn’t make it to Pulse 2012 this year, but we sent Tom and Donnie over to check out what IBM Tivoli has to say about “Optimising The World’s Infrastructure”.

IBM streamed the keynotes so I was able to keep abreast of some of the content, and it seems like IBM Software is finally nailing it when it comes to customer references at events speaking fluently about business issues and best practices. Pulse began to get this right in 2009.

Given I couldn’t make the event, what could I really add? Well plenty obviously- the key themes of mobility, analytics and virtualisation are all worth talking to in a management context – take IT security, which now needs to both encompass these themes, and be enabled by them. The beta release of SmartCloud Continuous Deployment server is another big deal.

But as someone growing into the “elder” role I increasingly see it as important to remind everyone of how we got here.

Tivoli went public in 1995. IBM acquired the company outright for $750m+ in 1996, creating today’s thriving Austin Angel and VC scene in the process. At the time IBM was instituionally still pretty clueless about running an independent software business.

Tivoli’s CEO Frank Moss as the time described the deal as a “reverse takeover of IBM by Tivoli”. It was pretty easy to scoff at the time, but Tivoli changed the culture at IBM permanently and fundamentally. Arguably the Tivoli acquisition was the one that really set IBM on the way to its current Software and Services success.

It starts with developers. Well it would, wouldn’t it? Tivoli engineers were used to developing on Sun workstations. They were developing Unix software, and Sun built the best workstations at the time. Soon after the acquisition was announced Tivoli bought some new hardware for development (well you would, wouldn’t you?) or rather it tried to. Sure enough an IBM bean counter popped up and tried to scotch the deal. IBM of course sold its own RS/6000 workstations, and of course Tivoli developers should be using those instead. The absurdities of IBM internal purchasing aside, this was an existential moment for Tivoli, and in fact for IBM’s entire future. Tivoli fought back and insisted that the developers got their way- I don’t whether Martin Neath, Tom Bishop or Frank Moss deserves the credit, but the end result was that Tivoli could keep building Open Systems software, using Open Systems hardware- but it could choose suppliers. Why was this event such a big deal? The answer is pretty simple – if Tivoli was perceived to be built on, and optimised for, IBM hardware then IBM had thrown away its money – you don’t buy a cross platform management framework while Unix is exploding and then hobble the acquisition.

Solaris may be a falling knife today but it remains one of IBM Software’s key customer platforms (indeed for a long time it was IBM Software’s biggest market for WebSphere sales. And so the business grew. Sometime in late 1999 it became apparent to me that IBM was actually serious about this software stuff). I remember turning to my boss at the time, Jonathan Eunice, and saying exactly that. We’d just met the Software leadership team and been impressed.

But then in May 2001 came another existential challenge for Tivoli – IBM Systems and Technology Group announced project eLiza – with the goal that IBM hardware would be self-managing and autonomic. You wouldn’t need to buy management third party software because it came built in. If Tivoli wasn’t going to optimise for IBM Hardware, then IBM Hardware was going to build its own management tooling. It my opinion this was the point at which Software Group really took charge. Robert LeBlanc called a summit, supported from above by Steve Mills, and below by Sandy Carter, to establish a common position on Tivoli, Software broadly, and how IBM Hardware and Software would work together. In short, Tivoli won. Software won. This is not to say that STG didn’t continue to build management tools- see System Director – but just that Tivoli was going to be part of the mix, too.

Software and Hardware started to become less adversarial, more collegiate, more matrix-structured and capability, rather than brand, driven. The watch words and approaches of IBM Software today were already extent back then. And so in recent times it came to pass that Software Group took over Systems and Technology Group (See my coverage here).

Arguably we have come full circle, or at least seen the pendulum swing away from Open Systems. IBM is beating a drum called Smarter Computing, where IBM (and partner) Software run optimally on IBM hardware. But I believe that without the lessons Tivoli taught IBM, we wouldn’t be seeing IBM’s share price where it is today. Lou Gerstner was all all about integrating customers systems, and Tivoli fit that bill perfectly. Gerstner set the direction, Mills executed, and Tivoli was a critical delivery vehicle.

Finally though – I had to chuckle at some of the Pulse speeches, which kept talking a new mission breaking down management silos, managing everything holistically across organisational and platform boundaries. This new mission was in fact Tivoli’s mission back in 1990. When Moss talked to a management Camelot in speeches just before the deal, he was already talking integration with building management systems and non-traditional IT networks such as manufacturing.

The Tivoli vision has legs, and I am only now beginning to see just how smart the IBM team was that did the deal, and how long its planning horizons were. IBM executives don’t just play quarter by quarter – they play for the long term.

disclosure: IBM is a client.