Network Automation: How companies are making the transition to a SaaS model

Dustin Snell was the founder and CEO of business process automation software company Network Automation, and then took on the role of CTO. Now, after two years, he is back as CEO.

Based in Los Angeles, California, Network Automation has just finished its eighth year with a record quarter, up 47 percent from the previous year's Q4, during which it also added 68 new customers. Nestlé Waters, BMW USA, and Payless Shoe Source complemented its existing customers, which already included Starbucks, Kaiser Permanente, Verizon, and Whole Foods. It is focused on enabling these companies to build a wide variety of different solutions without the need to write code, using its AutoMate platform.

To set the background, why have automation companies become such a hot commodity lately?.

There are some key drivers making automation more relevant. In the same way the web gives people access to a wealth of information, the rise of the cloud and virtualization has given companies access to practically infinite business resources. Often, you can't get enough personnel fast enough to harness and manage those resources, so we provide automation to help integrate and manage them effectively. Whether you see it as expanded reach or technology sprawl, automation has to be a part of that picture.

As far as what is making automation particularly hot right now, it's the exponential growth of computing assets and resources. For our company specifically, introducing new features, improving stability, listening to our customers, and constantly refining and honing our marketing message are all affecting our results. There is also a lot of activity among larger companies, which I see benefiting us in getting the message out there and growing the pie. The level of support in the analyst community of the need for automation has picked up too, as opposed to just relying on scripting, job scheduling, or application integration.

You recently moved back from the CTO/Software Architect role to reassume the helm as CEO. How has that change affected your perspective?

To be able to get outside the CEO role was an invaluable exercise; to be outside the fishbowl. What your customers tell you is very valuable, and we pay attention to that. But I think it is just as important to go and listen to some of the people who are not yet your customers. Leaving the CEO role and also doing some consulting during that two years gave me a unique perspective on what's happening in the world with businesses, the needs they have, and the challenges they face. So I've come back to that role with a fresh perspective and strategies to acquire some of those prospects as new customers.

"If you're big enough to have an IT department, you're big enough to need automation software."

—Dustin Snell, CEO of Network Automation

We're in a fortuitous position now because our technology is mature. I've always been a technical CEO, closer to a Bill Gates than a Steve Ballmer. We think automation is for everyone. If you're big enough to have an IT department, you're big enough to need automation software. However, we are not exclusively working with enterprise. Our software centers around 500 available actions that span various processes, protocols, applications, file transfer, and reading and writing from databases. Those are all business processes that include IT processes as well. About 80 percent of our users are IT professionals, but the other 20 percent are solely business professionals, and it enables them to build solutions without writing any code; just put it into production and automate it. Very often, they are application integration solutions, or they can be just automating a business process.

This year, we've seen the number of mergers and acquisitions go down overall, but SaaS and cloud companies are still a hot commodity at 15 percent of the global M&A total. How has this trend changed the playing field?

I think opportunity is always created when larger, less nimble players take out smaller innovation factories. A few years ago, when Microsoft bought Opalis, which was one of our primary competitors, it resulted in a windfall of business for us. We had companies coming to us who wanted the level of responsiveness a smaller company could provide. Pitney Bowes was one of those. The Opalis technology ended up fading into some more obscure part of SystemCenter. They use it in more of a vertical context that makes sense for Microsoft. At the time, I remember we thought that deal might be threatening us with Microsoft getting into our space, but in practice, it didn't work that way. It turned out to be a significant opportunity for us.

Do you see Salesforce as a bellwether for SaaS valuation?

First of all, my roots are in CRM. In the 1990s, I worked on the development team of Goldmine. I believe CRM is one of those applications that lends itself very well to SaaS, given that it is primarily a productivity application for sales and customer service. There are also many players out there trying very hard to fit a square peg in a round hole. SaaS isn't the ideal for every company. I see subscription models being entirely separate from SaaS. A subscription model can apply to almost any business software when it needs to scale, and where running through a web browser doesn't necessarily apply as broadly. So it depends on your definition of Software as a Service. Some would argue that it has to have all the elements of not being installed, running through a browser, hosted in the cloud, and subscription-based.

In terms of revenue growth though, as in the nice smooth revenue growth that Salesforce has, it's their subscription model. They do a great job with their software, and they pay very close attention to their attrition rate. For integration and automation, it's going to be much more often on premise. As more organizations have more of their assets in the cloud, it will make more sense for an integration and automation platform to be in the cloud too. It's not there yet. Right now there are some startups that have the majority of their assets in the cloud, but most companies still have them on premise.

So what is your expectation with regard to that kind of movement?

They are moving to the cloud, and they want a cloud-aware system. But as far as living in the cloud, just think about which is more challenging; to have software reach from the cloud through your firewall into your organization, or living in your organization and reaching out into the cloud as needed. Our software does not live in the cloud yet, although it's definitely on our roadmap. Right now, we are focused on adding cloud-related actions like Amazon Web Services, Windows Azure and Office 365 services that are in the cloud and we can talk to, but where we can integrate on-premise in a way that makes sense.

With our Automate 10 version this year, we're going to be introducing subscription pricing, which is something normally associated with SaaS. I would say subscription is the aspect from a revenue growth perspective that will give you those nice smooth indicators. If you can get your entry price point low, your attrition rate low, and payments being made monthly or quarterly, you can have all the benefits of SaaS without necessarily running through a web browser.

Realistically, how long do you expect that kind of transition to take?

For us, the technology to go forward with a subscription model is going to be available about the middle of the year. How it gets turned on and transitioned from a conventional license model to a SaaS-type subscription model is a process where any software business needs to pay close attention. It represents a complete change in your business model if you think about it. If 100 percent of our existing customers decide they would rather pay monthly, and we don't gain any new customers, what does that do to our cash flow?

We're going to be moving forward carefully, with the eventual goal of opening up the door to our technology to many more users at a price point for those who would prefer to put it on their expense account and treat it like a utility. But it doesn't have to run through a web browser to do that for a technology like ours. For CRM software, it's there now and it works great. Eloqua is a good example. We've just signed up with them, and we're sending people out to get trained and start using it. But for automation and integration software, it's still not there yet to run through a web browser or even in a cloud stack.

Has your view on the business-technology relationship changed?

I'm more of a technology-oriented person. Accounting is not my favorite subject. When I was CTO, I was primarily focused on our technology roadmap and evangelizing the product vision. As CEO, I have much broader, more holistic responsibilities related to the entire business, which has forced me to look at issues like accounting, which are equally critical.

The way we recognize revenue is a good example. Right now we're doing more service center business, and have more recurring revenue maintenance dollars, and subscriptions are going to be a big part of that. We are privately held, so for us the requirements are not as stringent as a publicly traded company, but these factors still need to be given consideration. Revenue flow is just one of the things we have to juggle as we look at demand for new business models like subscription.

How do marketplace clearinghouses specifically for SaaS applications in business-like AppDirect and SaaS Markets fit into the picture?

There are a couple things these marketplaces enable, especially for lower-dollar value transactions, like purchasing software with a few clicks. There is a reason e-commerce has never really caught on for $50,000 transactions though. As we go to a subscription model, it will lower our entry point and make it more applicable to us.

For higher amounts, people want to have more process with different levels of approval, but there are other elements of the technology like ease of installation that are easy to appreciate. Look at Apple's AppStore, which is installing native software of course, but it catalogs it for you, makes it easier to remove or upgrade, manage centrally, or reinstall when you get a new device. Those are all advantages regardless of price points.

We don't have a spot in the AppStore yet, but we're working right now on getting listed in the Windows store. Also, if we have a Salesforce integration action in our product, we would love to be featured in their store under automation products that integrate with their platform.

How do you keep up with the demands of your customer expectations in terms of SLAs?

Right now it's different from a guaranteed uptime model, which is what we will need to consider in making the transition to SaaS. What our customers expect is that we back up our product with redundancy, failover and on-premise checkups. When we do an SLA, those are the kinds of issues we are dealing with, as opposed to five-nines uptime. Our sales engineers do consultative services including regular check-up services to ensure that everything is working properly. We write those standards into our SLAs, as well as failover provisions, all of which result in an assurance of uptime. We just don't express it in those terms.

"We are pursuing a middle path between PaaS and SaaS; between commercial software and building your own."

— Dustin Snell

How important is the role of Platform as a Service?

PaaS is extremely important. The difference between a pre-built application and a custom one is speed versus flexibility. Without PaaS, the near infinite research that's available in the cloud would only be available to pre-built commercial applications like Salesforce.

Sometimes, because of legacy considerations or specific business needs, companies will have a need to create their own applications. PaaS allows those applications to scale in the same way a pre-built commercial application would. We are pursuing a middle path between PaaS and SaaS; between commercial software and building your own. That's the unique place in the market we want to occupy, with an application integration platform that doesn't really do anything out of the box like a pre-packaged application does. It also doesn't require you to write code like a pure platform would. So it's far faster to implement than writing your own system, but it maintains the flexibility of a custom solution. We say it exists between software and a platform. Whether it should live in the cloud or not depends on the proximity of the system to the assets it integrates.

For most organizations, that is still primarily on premise, but with the cloud increasingly involved in the mix, it's kind of a hybrid scenario. So PaaS is still critically important. I would hate to see a world of pure SaaS where you lose that flexibility to create your own solution and have access to the native platform.

Different businesses have different needs and some want to be closer to the layer, with fewer abstractions from the platform they are working with in order to maximize flexibility. It increases time-to-market the closer you get to the bare metal, but it gives you more flexibility. That's what we are all about right now is riding that line between the two.

Are you considering working with cloud service brokers to help with the management of things like identity provisioning, security or service level management?

As we get closer on our roadmap to pure cloud play, we will definitely be considering brokers. We've aligned with an inflection point when the businesses out there buying our technology have a majority of their computing assets in the cloud. We'll be ready with a pure cloud play when that happens.

We recognize SaaS is an increasing part of the application matrix for businesses. What we do is help people automate applications of any kind, whether SaaS or on-premise. We have large parts of our software completely devoted to cloud technologies. We live on-premise right now, but we are entirely SaaS-enabled. We can definitely talk to SaaS applications, as well as the entire AWS stack, and the Windows and Doors stack on the path side, and integrate your business processes with your automation solutions.

We are taking the components of SaaS that work for us right now, which are first talking to SaaS applications and bringing them into the mix, and offering our software on a subscription basis. But we are positioned to move to something fully hosted in the cloud as soon as the demand is there to justify it.

Dustin Snell started Network Automation in 2004, and has grown the company to more than 10,000 clients across 60 countries to build solutions covering job scheduling, automated FTP, batch processing, automated backups, scripting, and event log monitoring, to name a few. More information and a 30-day free trial of the AutoMate platform is available at www.networkautomation.com, or you can follow on twitter @NetAuto.