Interview: Doug Burgum on Microsoft's business apps plan

MBS chief emphasizes gradual evolution and migration

Microsoft spooked business applications customers two years ago with talk of an ambitious strategy to roll together its disparate Microsoft Business Solutions (MBS) software lines into one converged code base, an initiative called "Project Green." At its Convergence business applications conference this week, MBS head Doug Burgum assured attendees that MBS' five main product lines have years of life remaining.

He mapped MBS strategy in a keynote speech delivered 22 years to the day after he started working at Great Plains. MBS was created several years ago when Microsoft combined business applications development with acquisitions of midmarket ERP (enterprise resource planning) software makers Great Plains and Navision. Although Microsoft owns the desktop, it's among thousands of vendors vying for market share in applications to manage corporate operations such as sales, human resources, customer service and accounting. Burgum's task is to raise Microsoft's profile.

After his keynote Monday, Burgum met with IDG News Service to discuss the fading of Green and MBS' strategy to increase its market heft. Following is an edited transcript of the interview.

IDGNS: Microsoft is talking about Project Green in a different way from two years ago, with it being less of a new code base and more of a gradual, incremental process. What prompted that change?

Burgum: We haven't changed the endpoint. The vision remains the same. It's a little more crisp because we've had two more years of learning, and because of the deep customer research that we did.

There are two elements that led to a different strategy on how we'd like to get there. One is related to learning, both from our customer base and our partner base, that if the innovation could be delivered in ways that are more digestible and more evolutionary, that that would be preferable to dramatic, revolutionary change. So, some of it was to make sure we were pacing it right. The other piece was synching up with the refined Longhorn schedule.

IDGNS: Are there specific elements of the Longhorn architecture that will enable "wave two," planned for 2008 and beyond, with the business applications?

Burgum: Yes. An example there would be WinFS, which is the new file system. (Microsoft removed WinFS from Longhorn last year and said it will be released later.) ERP systems have been more around structured data, and the way people work is around structured and unstructured. With WinFS, it will be easier for us to light up that capability.

If a person is doing order entry work, and there's a dispute with a customer over an invoice, what happens today is lots of e-mails go back and forth with the customer. It's rough to be able to easily link those, do a search on them, all those things. But if you could put the unstructured data, the e-mail communication, with the structured data -- here's the invoice, here's the accounting entry that led to that invoice -- having the structured and the unstructured together, we think, would have a lot of value to customers. That's the kind of thing that you'll see more in wave two.

IDGNS: The word that's been used a lot today is migration. The idea now is that customers will be able to get to the eventual converged code base, to Project Green, with upgrades on their individual product lines. There will never be a rip-and-replace?

Burgum: Exactly. Over time you'll see shared design. There will be more shared design in wave one. For instance, we're going to have 50 roles in each of the ERP products, based on 50 roles from the customer models we have. We will have a role-based user interface. If I'm an AP (accounts payable) clerk, when I log on I get my AP clerk interface. So that's shared design, but in wave one, it won't necessarily be shared code. As we move forward, more of the shared design will be the shared code. As that converges, the percentage of shared code increases, and it becomes a more evolutionary than revolutionary change.

IDGNS: We've heard from a lot of people using Microsoft CRM (customer relationship management) that the deep integration they're looking for between CRM and Outlook and Office isn't there yet. It sounds like that's on track to happen in the next release, due at the end of 2005. By then, it will have been three years since CRM 1.0. Why such a long development path?

Burgum: During that timeframe there was a 1.2 release that probably should have been called 1.5. It makes it sound like a point release, when in fact we did localization for a dozen geographies and added a lot of capabilities. In my mind, 2.0 is really more like a 3.0 release.

Obviously, we hadn't intended to have this much distance between releases, since 2.0 is coming later than we had hoped. We're making sure that we get it right. It's never easy to announce a delay in a release, but it was the right thing to do.

IDGNS: What led to the delay?

Burgum: It's a combination of things taking longer than what was expected and us being more aggressive about the integration scenarios and some of the functionality. The choice would have been (to) release it on time and cut features. We decided we didn't want to cut features. We'd rather keep features, add some features, and bend the release date.

IGDNS: Analysts have said the 2.0 version is where Microsoft CRM will be competitive with other products out there. If that's the case, are you comfortable with having released CRM 1.0 more than two years ago?

Burgum: Yes. Relative to any other midmarket CRM product, it's been the fastest to 2,000 customers, the fastest to 3,000 customers. We have some very successful (independent software vendors); we have a lot of happy customers. We've been able to have real market feedback -- you can't match any level of research you do with having a product in the market.

IDGNS: Last year around this time is when Microsoft CFO John Connors said on Microsoft's earnings call that MBS wasn't living up to expectations. Last quarter saw flat revenue. What's the strategy for addressing those issues?

Burgum: We had flat revenue against 41 percent growth a year earlier. We also had nice, healthy increases in license revenue and customer additions. We don't release those numbers, so nobody gets to see them, but from a health index, we feel good about that.

What led to the flat revenue in part was the decline in services revenue related to a change in the business model with partners. We lost a percentage of the revenue and that disrupted the services feed. And then also, we got the tail end of our exiting of the online advertising strategy for bCentral. The declines are planned declines because we are trying to lead more of our services business through our partners, and we're not in the advertising business. Our sense is that we're still outgrowing the market. We feel pretty good about where we're at.

IDGNS: There were issues with the channel last year, consolidating to one partner program and adding MBS, and with moving CRM into the volume licensing program. Is that settled now?

Burgum: My sense is yes. There still is more upside for us for to get our execution even tighter, to get a second year under our belt with executing against the new program and the new structure. But by and large, the feedback is good. Our partner satisfaction numbers are back up again -- they had gone down during all the changes. The feedback that I've received from partners is that they all feel like the field execution is quite strong. They feel like they've got good relationships with their partner account managers, they’re more tied into the go-to-market strategy. I think everybody is feeling like we made it through that transition.

IDGNS: One thing you talked about a lot today is verticalization. What is Microsoft's strategy on that? Do you want to see the verticalization coming from the partner channel? Would you develop vertical versions of applications made available directly?

Burgum: We really see the vertical strategy as being partner-led. There are cases where we think adding functionality helps, like the case of Great Plains not-for-profit, where we acquired some functionality and added that to the core base. People have gotten confused about that, thinking we're trying to go all the way vertical, when in fact even our ISVs were saying, "Hey, it's actually better if you have this core functionality." We really do believe the vertical effort will be partner-led, but we want to be a great foundation.

IDGNS: So you don't see buying something like supply-chain software maker Manhattan Associates, which is a partner?

Burgum: From an acquisitions standpoint, other than giving you a "no comment," I wouldn't make a blanket statement excluding any possibility if we felt that it helped support the strategy. You saw SAP bought Retek last week -- that's not what we'd be interested in. If products are sold through a direct sales force at high prices, we're probably not interested. (Oracle later made a counteroffer for Retek, higher than SAP's.)

Copyright © 2005 IDG Communications, Inc.

How to choose a low-code development platform