Deep Blue veteran plots Lotus game-plan

As the man who decided in 1996 that IBM should pit its Deep Blue supercomputer against the then chess world champion Garry Kasparov, Ambouj Goyal is used to thinking strategically. Now charged with heading up IBM's Lotus software division, Goyal is going to need all his powers of perception to go head to head with Microsoft in the play for domination of the collaboration-software market.

Since IBM acquired Lotus in 1995, it has been trying to integrate the company's software with its own Websphere J2EE middleware platform. IBM sees it as natural fit and argues there is no difference between how people interact with a portal and what they do with collaboration software.

The marriage between Websphere and Lotus is being integrated under the umbrella of IBM's new Workplace platform – which is an overall plan to add J2EE Web services functionality to various IBM productivity tools.

But not everyone is on-message when it comes to the intertwined future of Websphere and Lotus. Analyst firm The Radicati Group recently produced a report claiming that, although innovative, the new Workplace strategy will actually damage rather than rejuvenate Lotus. The organisation claims that Lotus customers are concerned with the 'momentous task of migrating Lotus Domino data and applications to the J2EE-based Workplace platform'.

ZDNet caught up with Dr Goyal to find out which is the bigger challenge: convincing customers that Workplace rather than Exchange is the future of messaging, or taking on a grand master at his own game.

What did you make of the Radacti report? Do you think that they have a point about the whole Workplace strategy and the upgrade path for Notes/Domino being rather ambiguous at the moment?
Their conclusions don't make sense to me at all. Radicati are one of the companies that six or eight months ago reported the exact opposite of what they are saying now and the input was exactly the same. If the input remains the same and their output changes dramatically, then I have a basic problem with their conclusions.

Why do you think they are so way-out on their numbers, particularly regarding your steady loss of market share to Microsoft Exchange -- which they claim is a much more focused product?
Ask them why did they change? I announced at Lotusphere that over 1500 customers migrated from competitive platforms last year.

With server consolidation, customers are typically moving from multiple small servers to one or two larger machines. If you are a caught in a box and say 'whoops, I can't do it because my environment runs only in Windows', then you slow down the server consolidation the customer wants to do. Since our Domino environment runs across multiple servers -- you can choose Windows, Unix, Linux, AIX or even mainframe -- when customers are doing server consolidation they are doing Domino. Server consolidation is one of the reasons why customers are switching, the next is platform choice and flexibility, and the third is security.

So those are the three or four reasons why customers are migrating to our platform and the conclusions that are reached in this report are inconsistent with the realities of the world.

What about their figures claiming that IBM's combined market share will fall from 24 percent to 17 percent over the next four years while Microsoft will experience steady growth to 33 percent share by 2008?
I don't know how they came up with that. In IBM we don't report brand-orientated revenue but after every quarter our CFO reports significant progress in a vertical environment and he chose to report that our base messaging business grew by double digits. So how did we grow double-digit percentages in a declining market? IDC just reported numbers and in 2003 they have us with a bigger market share than Microsoft does in the same market. We have never lost our No.1 position since our acquisition of Lotus in 1995.

What about the general marketing of Workplace -- there is some criticism that you haven't got the message over about Workplace adequately.
We have to make customers happy and give them choice and flexibility. Microsoft gives no choice. I think a choice-based strategy has a better chance to gain share than a strategy that says: 'You must buy my server, my operating system, my directory, my database, my desktop -- you must buy everything we provide or you are a second class citizen, you're ostracised.

But you are confident that you are getting the Workplace message through to customers?
I think the communication eventually gets through when people get comfortable. When I came on board 15 months ago, nine out of the 10 questions that were being asked were -- so what is the migration strategy? Today it is only one out of the 10 questions when I visit the customers. It was the same thing when we first shipped Linux, people were saying 'so, is your AIX dead?'.

How much do people really need a function rich, collaborative email platform such as Workplace given the general move toward outsourced applications? Isn’t it simpler to just outsource the email problem?
By 1990 the outsourcing market was a few million dollars, it is now a few hundred million. There is a general trend towards outsourcing. But email is the same as any business-critical business application and you have to really consider what advantage you will get from outsourcing it.

You have been at IBM for over 20 years now and were around when the company seemed to lose its way in the early to mid 90s. There seems to be a general paranoia in Microsoft to avoid making the same mistakes IBM did -- can you see any parallels in Microsoft's behaviour now and IBM 10 years ago?
Every company goes through this. You want to protect your existing products so much that you don't take advantage of disruptive technology.

One hundred per cent of Microsoft's profits come from the desktop operating system, the rest of them combined lose money. So they have to force an upgrade cycle every few years, and if they don't, where is their profit going to come out? To force the upgrade they have to build Longhorn -- which I call a personal supercomputer: two-way SMP, 16GB of memory, 100GB of disk drive. And they have to upgrade the masses, hundreds of millions of people onto this personal supercomputer -- if they don’t, their stock-price is hurt.

This is a dilemma. If someone does provide technology for the masses, that is good enough, that's very disruptive. Microsoft has to avoid that destructive move and the only way they can do that is to get rid of their cash engine. Just like IBM decided that no matter what the question, the answer is mainframe, Microsoft has decided no matter what the question, the answer is Longhorn.

I understand that you were involved in the Deep Blue project?
Yes, those were the days.

What was your part in Deep Blue?
I was the manager of the project. My key role in that is that I killed the project. The point was that they were doing research for eight years and my view was that either they make the big play or they don't play at all. They asked me what was the big play and I said you have to beat the reigning champion.

So going up against Garry Kasparov was your idea?
I didn't do anything with the technology, I just pushed them and said there is no point in spending money unless we can really show that we can play the big game.

Did you manage to do see the film about the match -- Game Over -- which came out last year?
I have read some books on it but haven't seen the movie

What's your take on the various conspiracy theories around the game -- that Deep Blue was being manipulated by a human for instance? And that the machine was dismantled straight away without giving Kasparov the opportunity of a rematch?
The whole idea for doing this is to see if the technology can reach a point where we can pass a threshold that has been a 100-year-old computing problem. When the threshold was reached, there was no reason for me as the head of computer science to continue to invest. From an IBM research perspective, we did our job, many of the technologies developed from that time ended up in our supercomputer systems.

The director of the film -- Vikram Jayanti -- made some comments that IBM was under a lot of pressure to win that contest and would have gone to any lengths to ensure victory. Was there a lot of pressure from the rest of the company to win?
No, the pressure was from me on my team because we were trying to solve a scientific quest. It was an important problem -- similar to the first time a machine went faster than a human being; so could a machine beat a reigning chess champion was a big question.

In 1993 when Lou Gerstner came along we weren't doing particularly well, and there was a general feeling that IBM was a dinosaur. In that environment anything that came out from IBM that showed we were investing in innovation said 'IBM isn't a stodgy company and can actually deliver world-class technology'.

What is it about the culture at IBM that has kept you there for so many years?
I have found IBM to be largest sand-box that I have ever come across. They are prepared to invest in industry-transforming technology.

It's interesting that IBM have put someone of your technical pedigree in charge of Lotus rather than someone with less technical expertise but more focus on marketing and winning the numbers game with Microsoft.
Al Zollar had done a fine job in transforming Lotus and getting to point where I could step in. I think the company had got to a point where it needed someone with the technical vision to move it on to the next stage. A transition comes in the industry and you need different skills -- although I am not that bad at marketing.