Covering new ground in a new year

Winston Prather is going into a year which no other HP 3000 division manager has ever faced: the beginning of HP’s certain end of its relationship with the 3000 community. He faces a major mission in the next few months in deciding what HP will do to help the OpenMPE movement, at the same time taking on the task of selling customers a technology many have avoided for several decades: Unix.

We spoke with Prather just two weeks after his division announced the end of HP support for the system. Our 90-minute talk ranged from the rationale behind the a decision which surprised many in its timing, to Prather’s second general manager job at HP for the Telecommunications Systems Division (TSY), as well as exploring the vendor’s future contributions to the HP 3000 community.

It’s been a busy time, but you apparently have more than one job to do now. How long have you been general manager of TSY, and what is that division’s business mission? Is it complementary in any way to the HP 3000?

It’s been like nine months. A lot of the general managers and above have done this at HP. Hey, when your boss asks you to do something…

[The business] is related to the telecommunications infrastructure, so it’s a new area for me, but something I’ve always been interested in. At the moment I’d say it’s a bit stealth, because we haven’t announced any products, kind of a startup inside HP. It’s been around in a small way for a year and a half, and then once the business plan became a reality and HP was ready to put some money behind it, that’s when they asked me if I’d like to get involved. It’s designed as a hardware solution that rides underneath vertical telecommunications software.

It’s not complementary to the 3000 at all. It’s complementary in the sense that it’s servers, a combination of all sorts of different architectures. TSY is a vertical business segment, as opposed to the 3000’s more horizontal play. You could run into a telecommunications customer who was using an HP 3000. These servers are outside the datacenter.

Is the 3000’s marketing director Christine Martino also at TSY?

She is the marketing manager there. Although it hasn’t taken a lot of her time, because she’s been pretty busy with 3000 stuff.

Has working at TSY changed your job at CSY in terms of time spent on each position?

I spend less time on the 3000 business now than I did initially. It’s not just the TSY thing, so it’s not as big a change as you might be implying. I’ve done other things as well on top of my GM role at CSY. That kind of thing happens all the time. You’ve got your daytime job and additional HP task forces.

Has TSY been getting an influx of CSY employees in the past two months?

There has been some, but not really that many. I can count the technical people on one hand. I don’t anticipate it being any more. There were a couple of people interested and I thought they’d be a good fit technically. They applied for a job.

Do you think you’re going to go from being CSY manager to TSY manager full time?

I don’t see that happening. The 3000 business still requires quite a bit of my time. It’s also what I enjoy doing most.

Are you going to be the last general manager of CSY?

Gosh, I don’t know. Part of me wants to say ‘I hope so.’ But there’s a negative sound to that, too.

Is CSY’s performance now being measured on how many of its customers migrate to HP platforms?

What’s important for CSY is that we take care of customers, and make them successful. We do measure customer loyalty, and whether they want to move, or whether they don’t.

As hard as it for customers to understand, we didn’t make this announcement because it’s in our best interests. We did it because we think it’s important for them to know what’s going on, and the fact that they need to be planning to change. We just rolled the entire product line. It would have been better for us to wait a year, sell those, and then make the announcement.

Why make this announcement now, just a few months after all your new systems are finally shipping?

Once you conclude this is the right thing, I think you need to tell customers as soon as you know. If you delay, where does that put you on honesty and integrity scale? The feedback that I’ve gotten in the last couple of weeks from customers is they appreciate that.

A number of the larger deals that were in the process — they’re still buying them. Customers are buying because they need it. If you needed it yesterday, you still need it today. This is a strategic decision, not a tactical one.

Customers will choose conflicting responses to the announcement. Some will want to Migrate, others Homestead. How does HP handle the conflicts and help everyone?

I think there’s a fine balance between letting customers know what you recommend and enabling customers who choose to stay to be successful. I intend to [enable them], but I don’t want to do that in a way that makes people think we still think that’s a good way to go.

Here’s the bottom line: MPE will be around forever. And we want to help that. This is in no way HP trying to kill MPE. We will explore and look at all the different options to enable what I’d call the afterlife — or at least the after-HP life, beyond 2006.

Having said all that — and I wish this wasn’t true — I don’t think it’s in a customer’s best interest to count on the 3000 community and HP together to run your business. I’ve got CIOs calling me and saying ‘I bet my business on my platform strategy, so what should I do?’ I cannot confidently beyond 2006 it will be a safe platform to be running your business on. Because of that, all the things we do to enable the afterlife will not change our recommendation — that you move. If it’s not alright now with HP backing the ecosystem, why do you think it would be after that?

Are you then assuming the ecosystem has already had a 10-count, and can’t get up off the mat?

I wouldn’t say it quite that strongly. There’s no trend that would make me think things might change. I can’t imagine anything that can turn around the application trend. The lack of applications… is going to get worse over time. As much as we’ve tried to turn that around, it’s just not going to happen. They’re very good solutions in the verticals we’ve focused on, but once you get outside those, it’s just not in good shape.

What about 3000 tool vendors, the people needed by the majority of the installed base?

The strongest part of the ecosystem is the tool vendors, because they’re incredibly loyal. But even the tool vendors will provide tools that help customers transition. They will only continue to invest at the level that their business allows them. Customers are moving, have been and the rate is increasing. With our announcement it’s going to increase even more.

Some serious part of the 3000 community hasn’t bought much from HP in awhile. Why do you think so?

It speaks to the reliability of the solution. But it also speaks to how strategic the [3000] is. I assume they’re investing in new IT stuff. Why aren’t they using the 3000 for it? In general, our customers have been branching for a number of years to the use of HP-UX, Netservers and some competitors. Those who use competitors have left HP completely.

This data is more from the high-end customers than the small customers. It could be related to the fact that HP has much more touch with large customers. I have less visibility into the smaller customers.

Computerworld printed a quote from you estimating the 3000 installed base at “several thousand.” Did they get it right?

I was quite annoyed at that article. I didn’t say the installed base was that size, because I don’t give numbers. I referred to numbers I kept seeing printed in the past.

So the installed base is bigger than several thousand?

I would assume so. I wouldn’t be in business if it wasn’t.

Another part of that Computerworld story said HP told them the 3000 futures discussed at HP World were just suggestions. Is that what you presented there?

I don’t think we’ve ever talked about suggestions. We try to be as honest as we can with the customers, and tell them what we’re currently planning.

But I saw that brilliant slide you presented at HP World, showing how NT’s Blue Screen of Death wasn’t a thing you’d have to suffer in the 3000 world. Now NT is being offered as an alternative for the 3000, right?

Different customers will make different choices, depending on how mission-critical your 3000 has been. Do I anticipate the majority of customers will move to an NT environment? I don’t. But there’s definitely a class of customer that will. I exaggerated about NT in my pitch. It’s getting better every day. Relative to my expectations of an enterprise system, they’re not there yet.

Would you say the ecosystem includes the customers not making much of an investment lately toward HP, but keeping up on support agreements with HP and its partners?

Anybody who uses the 3000 or contributes to it is part of the ecosystem. The erosion at the application level leads to erosion at the customer level. Application providers want volume.

What happened to the plan to dominate selected key verticals with 3000 applications?

That works for awhile, but even those application providers continue to look outside the 3000.

How does this decision work for the smaller customer who doesn’t rely on application providers?

It lays out a timeframe for how long HP will be able to serve them. Unfortunately there aren’t enough customers like them committed to the 3000 for the longer term to put us in a position to deliver that service. There’s a contingent of HP 3000 users who need nothing from HP. They’re going to live on long beyond 2006.

You’ve had a couple of challenges now, with this and Y2K in your tenure. How’s it feel?

I’m glad it was me. To be honest, for a certain period of time, when I realized what I thought I’d have to do, part of me said ‘I don’t want to be the one to do it.’ On the other hand, I think it’s much more appropriate. The easy way would have been to do something else, and let someone else make this call. I’m just as committed to the 3000 business now as I have been forever. It’s totally appropriate for me to be here and make sure we do the right things from here on out. I intend to do that.

Would having support revenue flow directly to the division have made a difference?

Not at all. We’ve been measuring the business for years across the entire value chain. We manage the entire business as a whole, support and storage and services revenue. We try to manage for the whole business through our profit objectives.

Did you decide that doing the IA-64 port would be a no-profit project?

We did the estimate years ago. We looked at investing in a port that we couldn’t complete for a number of years. By the time we’d complete it, what’s the state of the ecosystem? Which applications and customers are going to be willing to do that transition? We summed it up and said it didn’t make sense. We didn’t do an ROI analysis of the port. The real issue is once you have that port complete, will any ISVs port?

Well, if you’d done IA-64 sooner, would the ISVs have followed?

Some of them would have. Not all of them — and we couldn’t get it done that quick. Even if we had IA-64 done today and shipping today, it doesn’t change the trend. It would make it worse for ISVs, because they’d have to support multiple binaries.

Did large customers determine the fate of the system for smaller ones?

The large customers have provided more input as to where we go. The makeup of our customer advisory council is not all large customers. It’s 50 to 60 percent large customers, and then we populated it with lots of other types of customers.

You didn’t talk about this decision specifically with that Customer Advisory Council. Why not tell them you were considering this?

We hadn’t really had an Advisory Council meeting since we started thinking we were going into this decision. The council doesn’t work like that. I don’t think it would be appropriate to share that level of concern. Having that discussion with the Advisory Council is irreversible. How do you have that discussion with them, and then change your mind?

Will you remove the clock speed governor on the A-Class systems to increase low-end sales in the final two years?

How we do it is irrelevant. All the customers should care about is how much they pay for how much performance. We intend to enhance the A-Class. We’ll be adding performance at the low end and the high end. Compared to the 9x7s, the A-Class is a compelling story today. And we’re selling a lot of them. We’ll bring the PA-8700 to both those lines.

Did the lack of business growth after Y2K help you recognize you weren’t going to be able to reverse the 3000 trends?

Y2K caused a bump in business, and it slowed down after than more than I thought it would.

How much impact did not having the N-Class ready until 2001, but announcing it in 2000, make in that lack of pickup? Can you see why people might not have purchased, waiting for the newer systems?

Yes. Anytime we pre-announce we’re doing something it slows down sales. We built that in before we announced it. You always want to wait for the latest. This is something marketing people plan all the time. I don’t think it affected the trend. It’s not a decision about how many sales we’re making.

How about the loss of revenues to CSY as a result of the illegal broker activities? Serious impact, or not?

I have a feeling it accelerated the decline in our revenues. That doesn’t affect the ecosystem as much as it affects HP. The customers still bought stuff — they just got it at a hot price.

People are talking about a hobbyist’s license for MPE source code. Is this a good first step for an OpenMPE?

I have no problem showing our source code to people from a hobbyist perspective. I’ve always been an advocate for sharing source code.

Would sharing source code hurt HP in any way?

It’s not obvious to me. I tend to think not. I tend to think that HP would not consider that harmful to us. Those customers who would stay beyond 2006 don’t buy anything from us anyway.

Is HP willing to allow MPE to move beyond the HP umbrella?

HP is willing to allow MPE to live on. I don’t know anyone who’s said differently.

People use Microsoft operating systems with HP hardware today. Do you think an OpenMPE, from a third-party entity, could keep people buying HP hardware?

Would people stay on and eventually buy some HP systems? Probably. Is it material, financially? I don’t think so. Would we invest to make that happen? Probably not. I don’t want to stop MPE from living beyond HP, but the return on investment wouldn’t be worth it for us.

How soon do you think have to make a decision about licensing MPE to parties outside HP?

I don’t feel the need to hurry, other than I know in the chat rooms there’s a lot of discussion about it. It comes back to my feeling that, yes, I want to enable afterlife. But it doesn’t change my recommendation. If I think the majority of my major accounts — and maybe some medium and small accounts — need to do something different than [use HP 3000s], then what’s the hurry? What’s the difference between announcing this type of enablement in January versus waiting six months?

Because every day you delay, HP’s infrastructure to assist an OpenMPE movement decays. Engineers are leaving. One them in your division has said the decision needed to be made soon. What about that?

You have to make sure everything that you need to do before turning over MPE has been done. I have no problem with that kind of urgency. We’re talking about the difference between making the decision a week from now versus six months from now.

I don’t feel there’s that much difference. I don’t think it’s a two-year effort to do what we’d need to do to implement some form of Open Source. But we’re still learning what that takes.

I want to make sure we do the right thing, and we do it right. Making a snap decision to do something has historically never worked out to be a good thing for me. I don’t want to say something that makes a customer think they don’t need to plan.

I’m trying to do what’s right, not satisfy a few people posting in the chat rooms. Maybe what they want us to say is the right thing. But until we’re confident of that, I don’t think it’s the right thing to do.

Are you ready to take the needs of much smaller customers into account on this decision?

Absolutely. But if you’re the customer who wants the one neck to choke, the huge corporation behind you, I think you’ll think twice about running your business on a platform not backed by a major vendor. Most of our big customers are in that camp. Huge corporations are not about to go with it.

On the other hand there are customers who don’t buy support at all. That end of the spectrum are the ones that would benefit most.

What about goodwill HP derives from helping them? Some say they will drop HP altogether because HP walked away from the 3000.

I think that’s a very emotional reaction. I’d like to think that because we’ve helped them over all these years, that means something. I’d also like to think a company that’s trying to be honest means something. I know we have to keep earning it. I don’t think our competition would be as forthcoming.

Some of the emotional reaction that has happened has bothered me. I hope once they spend a little time thinking about it, they realize we’re not abandoning them. It’s true, the platform won’t be around forever, at least not from us.

Can the 3000 community survive without HP’s participation? Will there be a 3000 community beyond HP?

Yes, absolutely. It’s all a matter of what you mean by survive. Will people use the HP 3000 beyond 2006? Absolutely. Look at MPE V and other proprietary platforms. I need mission-critical customers to understand my recommendation, even if they don’t end up on an HP platform. I owe to the companies using this platform to make sure their businesses are safe in the long run.

There will be some community that exists. It’s a matter of how the 3000 will be used, what level of mission criticality. I don’t think there’s anything HP could do to kill it, and we would never want to. From my perspective personally, gosh, I would never do anything to do that. I think I would make sure HP didn’t, either. HP would have no motivation to do that. We think the opposite: we want to do everything we can to help customers, even if it’s a small subset of them.