How Salesforce.com Is Winning Mindshare with SaaS: An Interview with Peter Coffee

Salesforce.com's Director of Platform Research offers his thoughts on his company running with the SaaS tag, as well as how virtualization and cloud computing are impacting business.

Virtualization Review Editor in Chief Bruce Hoard recently interviewed Peter Coffee, director of Platform Research at Salesforce.com Inc. They discussed a number of topics relating to Salesforce.com's role as a Software as a Service provider, and the virtues of cloud computing and virtualization in the company's success.

VR: Describe your job responsibilities.
Coffee: Four years ago we decided that we were going to expose the same platform on which we build our own applications to independent developers. My job came into existence at that time to begin engaging with an audience that had not been a high priority for us previously. We had always dealt with heads of business units, like the VP of sales, and offering a platform meant we needed to be initiating many more conversations with people at the CIO level and with comparable IT responsibilities. So that's my job.

I spend most of my time in conversations with CIOs finding out what they don't know that is true, debunking what they think they know that isn't true, and bringing back to headquarters information that CIOs need from us that will make us a first-tier choice for any application they might want to build. I want to be on the same playing field as any of the other mainstream enterprise IT platforms, and increasingly we find that in very substantial organizations, there's pretty much a three-player choice to be made between doing a Java app, a .NET app or a Force.com app. As long as I can be sitting at the table when that choice is being made, I'm very confident that the economics and the business advantages of doing a cloud-based application development and deployment will win out.

VR: Specifically, what do these CIOs want from you? What are the themes that you hear over and over again?
Coffee: Interestingly, three years ago, a CIO would tend to challenge me on whether we could do what we say we do. They wanted to be certain that we really offered flexibility, power and versatility for integration, and, of course, security and operational assurance. In the last year, I've noticed a dramatic change where now, instead of demanding that I prove that we have those capabilities, they're now seeking my help in selling the story upward to their board of directors, to their chief compliance officer and outward to their customers that using the cloud is not merely prudent, but actually advantageous to them. So what CIOs want from me is active and informed assistance in anticipating and handling the questions that they're going to have to answer as they try to carry a cloud initiative forward in their own organizations. And certainly, the one thing that's always going to be asked is, "Are you as secure as a datacenter that I operate myself?" To which, increasingly, I'm able to say with complete confidence, "Not merely as secure, but more secure—and, even more important, more governable and more auditable than traditional IT environments have been. We can be more precise about who has what privileges, and far more accurate and timely in telling you who is doing what with their privileges than the traditional IT environment is able to be."

VR: Do they feel isolated in any way because something's going on off of their premises?
Coffee: They assume initially that they will have to tolerate reduced visibility and control. Then we start showing them the facilities that we can make available to them in areas like our trust.salesforce.com Web site with the real-time monitoring of system status and system transaction volume and speed. Then they see the kind of dashboarding facilities that show them not merely gross indicators, like people logging in, but much more granular indicators, such as, "Are there reports that have been defined that no one's running because they don't seem to know they are there, and are there workflows that have been created that don't seem to have caught on with users?" They can now start to do much more targeted investment in training and education of their users instead of having to guess what they don't know. They can look at an adoption dashboard and see what the users do not recognize as being available to them. And so, pretty quickly, they stop thinking of this as being a remote and opaque facility, and start to realize it's a very well-instrumented service that provides them with tremendous information that enables them to focus their operations and improve their service-provider relationships with their internal customers.

VR: It sounds like you're providing them analytics-based information?
Coffee: Both raw data and analytics of that data, yes. Simple quantifying information, but also really good dashboarding-type facilities that they can quickly see—this is in green, this is in yellow and so on. We place adoption among the most important measures of whether a customer is achieving value from the service because our business model is based on ensuring that someone who has begun a subscription term today will feel they've gotten enough value that they will renew that subscription when the term is up. So on day one, we start by introducing them to what we call a customer success manager who becomes actively involved with making sure they have a rapid adoption process and a good feedback loop. It's a different relationship because we don't sell technology and then walk away and leave it to the customer to make it work. We sell a service and that means that the engagement is continual throughout the entire term of the service subscription.

VR: What are the advantages of cloud computing and Software as a Service during hard economic times such as those we're currently experiencing?
Coffee: It's been fascinating to see the response of the marketplace to the cloud model during these past two or three years. Actually, the first time I ever briefed CIOs with the message that they should be actively seeking to reduce their need for capital as part of their strategic IT initiatives was in 2007 when Bear Sterns collapsed. If you think about it, that was the first warning bell that capital availability might start to become a real limitation to people's ability to get things done. And then, of course, we went into the much larger meltdown in 2008. At that time, the market was regarding us in the same space as companies like Microsoft or SAP, but when we brought in record net new customers quarter after quarter all through 2009 and 2010, they realized that we weren't selling software. What we were selling was the ability to develop and deploy important new business systems without capital, up front and at a scale where a pilot project could be done very quickly with very low technical and budget risk. If it worked, it could be scaled up very rapidly, and if it didn't catch fire for whatever reason, it could be killed with zero sunk or ongoing cost. During these hard times, it has been tremendous to offer companies a way to accelerate through the tunnel and come out in the daylight with sticky gains in market share and meaningful new modes of engagement with not only their partners, but their supply chains as well. They've been able to develop in an environment where a traditional IT approach simply would not have been able to get the resources to do that kind of innovation.

VR: What challenges does Salesforce.com face when it comes to managing growth and making sure your customer service doesn't suffer in the process?
Coffee: When people ask me what it's like to work for our chairman and CEO Marc Benioff, I'll just say, "Mark is widely regarded as having been a visionary of the cloud, but really his more remarkable achievement is that he's built a corporate culture at Salesforce.com where everyone who works at the company understands the business that we're in and understands the service provider nature of that business." We get comments, for example, after every Dreamforce conference, that someone's been talking to a Salesforce.com employee and only at the end of the conversation do they discover they haven't been talking to an account executive, but someone who's an administrative assistant to one of the engineers. Regardless, they were still having the same kind of awareness of their needs and attention to what kind of experience they were having. That kind of depth in the organization means that when we bring new people in, they get it right away.

We tend to get a lot of people who are refugees from other technology companies who come in the door and say, "This is amazing. I've never been in an environment like this before where I don't have to apologize for the product and I don't have to make excuses for failure to deliver and I get to have conversations with customers that are entirely focused on what's working and what to do next." By way of comparison, a lot of energy in a typical technology provider is diverted away from getting things done, where we're able to focus on accomplishments, and you can see that operationally in the fact that year after year we do three major updates to the service that represent typically about 100 new features. Three times a year—that's a pace of delivery of new function to our customers that traditional software providers simply can't achieve. That's because they have the burden and distraction of having multiple legacy versions of the product that are out there in the field, and they have customers who defer adopting upgrades because of the disruption and the cost of introducing a major upgrade into a traditional on-premises IT environment. In contrast, our customers are able to accept the introduction of an upgrade with a couple of weeks of preview and checking a couple of boxes, and when people come in on Monday morning, the system is just doing new and better things. That's one of the reasons we can handle the growth—all of our energy is focused on moving forward rather than fighting a continual rearguard action against legacy product.

VR: In the eyes of your customers, is virtualization viewed as an under-the-covers technology?
Coffee: Almost all of our customers use us in combination with other assets. Our smaller and newer customers are what some people call serverless companies, but when we start talking about enterprise-sized companies, we're talking about firms that have massive investments in a decade, maybe two decades or more, of IT infrastructure. It's essential that we see ourselves in that context of the other assets that those companies are going to continue using. So we don't ignore something that's not cloud. We work with customers every day to say, "OK, how should you be managing what you have so that the cloud can add as much value to it as possible?" And virtualization is one of those techniques. I resist the label of "private cloud" for a virtualized datacenter. I have nothing at all against the idea of a company saying, "We're going to do a major datacenter modernization initiative that will include the adoption of virtualization for more flexible allocation of resources and better handling of variable workloads." That's great. That's terrific. I highly encourage people to do that because, among other things, when you do that, your entire IT portfolio starts to look like a constellation of services, some of which you run on hardware that you own, some of which you invoke from externally operated systems like ours. So I think it's terrific for people to do that. I just think they should be careful not to be fooled by thinking that using the same technology used by Salesforce.com somehow turns assets that you own into a private cloud. It doesn't. It turns them into a modernized, state-of-the-art, on-premises datacenter, but it's still capital that you have to own and it's still a software stack that you have to maintain. But that said, virtualization in principle and also in practice is a very useful lens through which to see the things you do with IT and start to see all of them as some species of either internal or external service.

VR: Describe your relationship with Amazon.com Inc.
Coffee: It's a complementary relationship. If you walk into the supermarket, there's a rice and beans aisle and there's a deli counter. They're both in the same supermarket. Amazon's virtual machines and massive online storage facilities are rice and beans; they're some large-scale tasks you're going to want to do. If I need to fix dinner for 100 people, I'm not going to buy it at the deli counter. If I need to do a very CPU-intensive process for a very short period of time, something like Amazon Web Services is a great way to do that, and a lot of Force.com applications call out to Amazon Web Services facilities to handle intense peaks of computational workload or very bulky volumes of relatively unstructured storage. Meanwhile, we're over there in the deli counter. If you need to build a business app with rich workflows, precise management of privileges, dynamic feed and update behaviors like Facebook in the context of what we call the chatter environment for collaboration, and you want to push the result out to an iPad, you can build that in Force.com. And you can do it in, at most, one-fifth the time and possibly as little as one-tenth the time that it would take to build the same kind of capability on your own in a relatively unstructured environment like a cloud infrastructure at a service provider. What we're selling is not the same thing that Amazon is selling, and every time that they get more capable, they make our stuff more valuable. And every time we get more sophisticated, we make more people interested in using the complementary facilities that Amazon provides.

VR: What lessons has Salesforce.com learned about virtualization and cloud computing that you can pass along to our readers?
Coffee: There are a couple. The first is that the cloud conversation very often begins as a discussion of cost reduction. If you go back to a company that's had a successful cloud adoption six months later, you'll find—and I'm basing this on some studies that have been done by some of the major systems integrators—that most of them say, "We spent way too much time on the cost justification and failed to realize what an enormous business opportunity upside this represented. If we'd been thinking about the value-creation opportunity and the business process acceleration opportunity that this represented, we would've done this a whole lot sooner." And so it's really important to ask yourself, is your goal to do your IT as cheaply as possible? Because really, when you think about it, it's obviously not. People make investments in IT because they have compelling return on investment. If you look statically at the IT capability you have today and say, "OK, now let's do it cheaper," that's overlooking the fact there's enormous, untapped return on investment opportunity out there in whole new models of user behavior.

Just look at the iPad: A year ago it essentially didn't exist, and now if the iPad were split off from Apple, it would be a major Fortune 500 company. You've got a whole new mode of interaction. A year ago, maybe one in five CEOs was thinking of deploying iPads as business tools, and now it's getting close to the point where three-fifths are saying they actively plan to do tablet deployment. Well, if you're still working with an old software model that really doesn't understand how to deliver capability out to that kind of device through an App Store kind of model for provisioning, well, then you're just not going to be participating in that marketplace.

So, when I talk to people about cloud today, a lot of the time I don't even use the word cloud. I talk about unambiguous benefits of social interaction, by which I don't mean recreational, but a system based on collaboration and conversation instead of being based on static artifacts like documents and files. I'm talking about conversation and workflow as the new organizing metaphor. That's what I mean by social. Mobile, because the mobile device is no longer some kind of lightweight viewer of your real app that runs on a computer. Your mobile device increasingly is the main point of interaction that you have with content and function in the enterprise. Openness to us means that every service has a way to exchange information and to trade requests and responses with every other service out there. If you look at most of our customers, they're using us in combination with things like Amazon Web Services, FedEx and UPS, Hoovers and Dow Jones. They're also using us with their ERP and financials, which are still running in their datacenters, so openness is critical. Our notion of openness is heterogeneity, flexibility, interoperability in ways that allow customers to be continually seeking the best value and the most strategic innovation opportunities.

VR: Do you find that CIOs these days are not so concerned with a hard-dollar ROI as much as they are with getting the right technology?
Coffee: You know, I think it might be the other way around. As recently as the 1990s, you could argue that the job of CIOs was to be stewards of the technology acquisition process and maintaining their technology portfolios. More and more now, I meet CIOs who don't feel that their job is defined by the technology that they acquire and manage—it's defined by the value they create for the business, and as such they are much more focused on measurable business outcomes. I might have had a conversation with a CIO four years ago where he would ask me questions about our uptime percentage, for example. Just last week I was having a conversation with a CIO who said, "Well, what can you tell me about measuring outcomes like up-sell rates and first-call resolution rates in the call center?" That's a different point of view. They're starting to understand, among other things, that the lifecycle of technology is accelerating quite rapidly. If you look at people like Qualcomm, for example, who are doing the chipsets in all these mobile devices that run Android, there's independent research now showing that the cycle time of designing and bringing to market new devices in that marketplace today has shrunk from something like 14 months to something like five months over the last two years.

CIOs are increasingly aware that spending a year or two evaluating and deploying a technology instead of getting something in that's going to be available, on time and on budget—and delivering value as quickly as possible—is not the best approach for them because a year or two from now, you may have turned the crank and replaced it.

The services model frees them to move at the speed of that accelerating marketplace instead of getting tied down to a particular portfolio of technology. They can start to think about their strategic portfolios being the abstract data types that they've developed and the process definitions that they've developed that really make the business work. I had one CIO say, "We don't think about apps anymore. Applications turn into silos. We think about workflows." We have 349 workflows running in the Salesforce.com environment, and some of those workflows have millions of dollars of revenue attached to them. But what they see on their diagrams on their whiteboards is not an array of boxes of hardware or licenses of software. What they see are functioning business processes where data flows through various points of adding value, and business outcomes are the measure that matters.

VR: There still seems to be a lot of confusion out there about what cloud computing and virtualization are.
Coffee: Sometimes I hear people say, "Oh, I know what cloud computing is. Cloud computing is when you use virtualization." You know, it's not. Virtualization's a very important enabling technology, but if all you do is adopt virtualization and you don't ask yourself what the business outcome is that you're trying to achieve, then all you actually wind up doing is adding one more layer of complexity and abstraction and one more relatively complex skill set to your existing IT operations. So you've got to be thinking about the "why" rather than the "how" to have virtualization turn out to be a strategic investment and not just a garnish of new technology on your existing practices.


Subscribe on YouTube