What it Takes to be a SaaS Provider

See my interview with SaaS veteran Mike Seckler, who recently joined the board of Apprenda, which provides a SaaS platform for delivering multitenant .NET applications. Mike was a co-founder of SaaS pioneer Employease and after its acquisition in 2006 by payroll provider ADP served as its division vice president.

Discover what's different about delivering SaaS compared to a conventional software business — and why Employease spent $10 million building infrastructure that's now available ready-made from an emerging breed of SaaS platform providers.
##CONTINUE##
PW: So Mike, we first spoke probably eight years ago when you were at a company called Employease, which was one of the pioneering SaaS vendors in the days when we hadn't even thought of the term 'software-as-a-service', and which subsequently was bought by ADP. I believe you left ADP recently. With that experience, and obviously your involvement with Apprenda, I guess you probably know better than most what makes a SaaS operation tick.

MS: Well, I mean, I don't know if I know better than most. I mean people are talking about SaaS every day. But I definitely had a set of experiences that gave me some opinions on that. Just to dig into that, when we started our company, Employease, way back in '96, software-as-a-service wasn't really talked about as software-as-a-service.

We went through kind of the ASP world and On Demand. But what we learned kind of really came with the tech downturn in 2001 — because the interest in delivering applications over the web started to build and you had folks porting legacy applications, putting web front-ends on them and hosting them, and calling that the equivalent of software-as-a-service; and then you had many software-as-a-service start-ups. And with the downturn, going into it — if you believed every press release — we might've had 25 competitors delivering software-as-a-service in the HR category, but by 2003, we really had none.

And I think part of the reason for that was that folks misunderstood what was required to deliver software-as-a-service. They thought it was just delivering the same applications they had always built but with a web front-end and with someone hosting the application. But in reality, it required really rethinking the whole business model of software and realizing that you were no longer a software company, you had to become a service business.

And what's interesting is that you, obviously, as one of the pioneers had to learn how to do all of that. But even now, there are very few companies around that help provide automation of some of the best practice that people have discovered in the intervening years.

Yeah, I think it's difficult, particularly for folks who grew up in the software industry. I mean, building a great application that looks good on a screen and can demo well in a boardroom is obviously still important.

But there's a whole 'nother set of requirements for building a successful SaaS business. You need to be able to implement well. You need to be able to deliver upgrades to hundreds or thousands of clients over a weekend. You need to be able to keep the system always up and running and running very fast. You need to be able to help your clients get data where it needs to go.

And software companies that were built around shipping CDs out, and working with third parties to implement, and delivering upgrades every 18 to 24 months — it really was a fundamental shift in the business model. And that's why I think we're still seeing very few folks have mastered all that it takes to be a successful SaaS business.

So partly, it's an operational thing, isn't it? It's actually having to not just deliver the software but actually keep it running as well.

Yeah, you need to embrace — to be a successful SAAS business - you need to embrace delivering the client experience. [You need to] realize that you can build the best application in the world — but if it's not always up and running, or if it takes 5 to 7 seconds between pages to load, or you can't get upgrades out to hundreds of clients without disrupting their operations, you won't be successful. So it requires very much of a different mindset — one that is the mindset of a service company.

And does it also change the way that you package and price the application; is that an important element of it?

Yeah, and I think each application is a different set of requirements that do that. At Employease it was per employee per month. A number of applications do it per user. Some other applications do it on a transaction basis.

What's exciting about some of the platform-as-a-service providers that are coming is that they're not just saying, hey, look we can handle some of the SaaS specific things like multi-tenancy and hosting independence. But we can also build in the infrastructure to handle a variety of business models, and pricing relationships so that independent software vendors who are good at building applications can tap into an infrastructure that handles all those SaaS business requirements.

Yes, and I think people often underestimate the extent of that infrastructure. I like to call it "Service Delivery Management" — that whole field of doing the packaging, the pricing, doing the provisioning, doing the customer services, doing the billing and making sure that people pay and ...

Oh, it's huge. I mean the role-based kind of provisioning — making sure each user only has access to the data that they need to see, even though you're hosting data across thousands of clients and within each client you could have a hundred users each with different access rights. And then the flexibility to handle different kinds of financial arrangements to satisfy clients of different sizes, with different sets of requirements. I mean, that's really what the best of the platform-as-a-service community could really handle. And not just provide the tool, but also bring the insight that a software vendor that was more of a traditional one in the past perhaps needs their hand held a little bit as they go through the process.

Right. And I think the other thing that you did a lot at Employease was the backend integration as well, which again needs to be delivered as a continuous service and delivered to all the customers at once rather than it being something that customers individually implement themselves.

That's exactly right. I mean part of the huge value we brought at Employease was that we were aggregating data across thousands of clients. And in the world of HR, when you terminate an employee it sounds kind of simple. Well, you need to get that employee off the medical plan, the dental plan, the 401(k) plan. You need to stop paying them, so it affects payroll. You need to trigger the COBRA notification process and a variety of other things. And in our world, we got to the point where somebody could click a button and then all of those backend data transfers just happened. And it was a huge part of our value proposition — that we didn't know we were going to have to do when we started — but because we architected our system in the right way, we could send data to United Healthcare or to ADP (who ultimately acquired us) through one set of infrastructure across thousands of clients.

And I think, actually, that's one of the positive benefits. We hear a lot of reasons why people don't want to do SAAS. They're worried about security; they're worried about control over their data. And [we hear] very little — I mean the industry probably does a bad job of communicating all of those benefits — of having the integrations up and running, having the vendor responsible for the application, and having the choice of packaging and pricing.

Yeah, but here's the thing, we had to do this before there was anyone who could help us, right. So Employease — on these things that are all kind of invisible and don't show up in a demo to a prospect [but] that are critical to being successful — we surely spent over $10 million on getting all those things right.

When I think about Platform-as-a-Service providers like Apprenda, what's exciting about what an Apprenda is doing is that they make available to a software developer that infrastructure, that expertise — all that intelligence and SAAS infrastructure, available on a variable per-user per-month style basis — that we had [to do] without a partner like Apprenda. We had to spend $10 million in fixed costs and it was part of the reason why it took us six, seven years to get to profitability at Employease and had to raise a lot of venture capital. I don't know that software vendors who want to build SAAS businesses today need to go through that same thing.

Right, and that's very interesting that we're now starting to see platforms coming out that do actually do some of that work rather than having the SAAS vendors having to reinvent the wheel every time. So hopefully that will provide opportunity for more SAAS vendors to flourish in the future.

-----------------------------
BY Phil Wainewright
Source：ebizQ

Phil Wainewright specializes in on-demand services (SaaS, PaaS and cloud), Web 2.0/Enterprise 2.0 and services architectures (SOA). He has closely observed the evolution of SaaS for more than a decade since founding the ASPnews [provide link] website in 1998. He now writes the influential ZDNet Software as Services blog and, as CEO of strategic consulting firm Procullux Ventures, advises ISVs and other industry participants on SaaS success strategies. He also founded the Loosely Coupled blog to track enterprise adoption of web services and is a member of the Enterprise Irregulars network.