Evaluating Tech Startups: The Risks And Rewards

By John Foley,  InformationWeek
Nov. 10, 2007

Tech startups are enthusiastic about the prospect of selling to businesses, and rightly so. Opsware, VMware, Salesforce.com--they're just a few recent examples of startups that hit it huge, whether through buyout, IPO, or organic growth. Venture capital firms are pouring money into promising early-stage tech companies--$1.1 billion went toward 187 software deals in the third quarter, according to PricewaterhouseCoopers and the National Venture Capital Association--and Web 2.0 has everyone thinking again about all the business possibilities on the Internet. The pieces are in place.

Except for one. IT organizations, the ones with the purse strings, treat startup vendors like they're radioactive. Blame it on tight IT budgets, a preference for doing business with fewer vendors, and a once-burned, forever-cautious attitude after Web 1.0. The tech industry's innovation engine is revving, but CIOs have a foot on the brake.

"After a while, you don't take risks anymore," the CIO of a Fortune 500 manufacturing company said in a meeting of peers a few months ago. "Your world is so complicated, you can't take those risks." InformationWeek Research bears out that thinking. In our survey of 150 senior business technology executives earlier this year, 74% described their companies' IT cultures as being moderate or conservative; only 26% called themselves aggressive.

That's a harsh assessment from someone who has struck pay dirt with two tech startups: Opsware, which Hewlett-Packard is in the process of acquiring for $1.6 billion, and Netscape, whose 1995 IPO was one of the biggest in corporate history and which sold to AOL for $4.2 billion in 1999. Andreessen is now on his third startup, social networking site Ning.com. So does Andreessen see IT organizations warming to startups? Hardly. Entrepreneurs who go knocking at the data center, he says, are met with "deadbolts, chains, shackles, boiling oil, pits full of sharpened steel spikes, iron maidens."

Ning offers service options for businesses, but Andreessen isn't going after the IT crowd this time around. "If businesses want to use it, that's great," he says. "If not, there are 1.3-plus billion consumers on the Internet who will."

In the face of such concerns, many startups are pushing ahead with new software, services, and appliances to businesses. IT pros may be fewer in number than the billion-plus Web users, but they make up for that in buying power. The average IT budget for an InformationWeek subscriber is $48.7 million.

	Zenoss' CEO Karpovich is taking on BMC, CA, HP, and IBM in systems monitoring Photo by David Deal

And, some of that money does in fact go to startups. Opsware managed to exceed a $100 million annual revenue run rate selling to customers such as EDS, General Electric, and JPMorgan. Aveksa, a startup specializing in data access governance software, recently landed health insurer Cigna. PeopleSoft founder Dave Duffield's new HR-as-a-service company, Workday, has signed 26,000-employee Chiquita Brands. "They're the 'try it out' kind of guys," says Duffield. (As these examples show, a startup's lineage counts. Aveksa, Opsware, and Workday were all founded by well-connected software industry veterans.)

Just last month, 2-year-old Zenoss revealed that IPTV and wireless equipment provider UTStarcom has begun using its open source IT management software, which competes with established products from BMC, CA, Hewlett-Packard, and IBM. With annual revenue of $2.7 billion, UTStarcom is the kind of blue chip account that tends to reassure other potential customers. Thousands of companies use a free version of Zenoss' software, but only 50 or so have signed to use its enterprise version, which lists at $50,000 for 500 managed nodes.

Zenoss gives its software away in hopes of hooking paying customers later with more features, support, and indemnity from any legal claims. That's one way startups get around bureaucratic roadblocks in the IT department--find a few early adopters among IT staff who champion their cause. "No CIO ever chose Linux in the early days. It just showed up," says Zenoss founder and CEO Bill Karpovich.

IT managers tend to use startups to fill gaps in their infrastructures or to make processes faster, better, or cheaper. Craig Shumard, chief information security officer with Cigna, uses startups to tighten security when he can't find what he needs from established vendors. "They're more nimble," Shumard says.

Cigna's need to manage data access more methodically caused it to take a look at new software from Aveksa for establishing user access privileges. Last year, when Aveksa first approached Cigna, the startup was only 2 years old, with few customers and not much of a track record. But Aveksa's executives were experts in identity management--CEO Deepak Taneja is the former CTO of Netegrity--and the company's software sounded like just what Shumard needed to put the finishing touches on a role-based access control system in development at Cigna since 2002.

The task was daunting--Cigna has 26,500 employees, 22,000 PCs, 9,000 laptops, thousands of applications, and 140 Tbytes of data. "I've got a lot of pressure on me from our business people looking to enable our business securely," Shumard says.

Aveksa's software promised to do something Cigna couldn't find elsewhere--let business managers collaborate with compliance personnel and security pros in defining roles that determine entitlements, or employees' data access rights. Promised because Aveksa's suite didn't fully support that capability at the time the company started talking about it. Aveksa's full-blown workflow app, Aveksa Role Manager, will become commercially available in December.

Cigna's experience underscores two advantages of working with startups: Early adopters can influence the tech road map, and they can steal a march on their competitors by getting functionality not generally available.

Vince Delperdang took the path less traveled two years ago when he chose Pillar Data Systems--at the time a relative newcomer to the data storage market--over EMC and HP. Delperdang, IT manager with O'Donnell/Atkins, a real estate and land development company in Irvine, Calif., used to stop by EMC's nearby office for lunch. But Pillar offered more storage for the price and better support with no services requirement--and won the business.

The Washington Post's Web site became an early user of Aggregate Knowledge's Web site recommendation software after WashingtonPost.com executive editor Jim Brady heard about Aggregate Knowledge last year when talking to a venture capitalist. Founded in 2005 and backed by $25 million in funding, Aggregate Knowledge's algorithms analyze what people do on a Web site, then make product and content recommendations based on what like-minded visitors have done.

Brady was impressed enough with the hosted service to ask the publisher's technical and financial teams to check it out. Such due diligence must be standard practice when working with startups for the first time. In addition to stress-testing the new technology, IT departments must know something about the company principals, check customer references, and even make an on-site visit to the vendor's offices (see story, "9 Questions To Ask A Tech Startup"). At Cigna, financial experts go over the startup's revenue, costs, debts, and assets, and they may get on the phone with the company's CFO or CEO for more information.

Even with those steps, however, there's an element of risk in turning over some aspect of your IT operation to an early-stage company. "On the Web, you try things knowing they may not work," Brady says. "If you're not failing a couple of times a year, you're not trying hard enough."

So far, so good with Aggregate Knowledge. The discovery engine makes content recommendations to WashingtonPost.com visitors, resulting in a stickier Web site. That's key to enterprise acceptance--a startup's ability to demonstrate that its technology delivers an advantage that can be measured in dollars, customer loyalty, or speed to market.

And if that technology can be implemented with minimal fuss, it stands a much better chance. The popularity of Salesforce and other on-demand software has as much to do with ease of implementation--no software to install, no servers to manage, no massive upgrades to deploy--as it does with ease of access for users. Software as a service can spur IT departments into action in response to user interest--or take the IT department out of the equation entirely.

The appeal to startups is obvious: They get to host their apps in Salesforce's data center and have access to its growing customer base. What's in it for IT departments? For one thing, they can spend less time vetting the scalability and security of the startup's on-demand facilities because Salesforce, having spent $100 million on redundant data centers over the past three years, already has passed muster. "You're piggybacking on all the work that we've done," says Ariel Kelman, Salesforce's senior director of platform product marketing.

The result is that big businesses are more likely to give the incubator's startups a chance. Right90 signed electronics company Sharp as a result of its relationship with Salesforce. Now Right90 has an influential reference account when it approaches other potential customers. "It's huge for us," says Paul Connolly, Right90's VP of business development. "You really need an enterprise-scale early adopter."

That's especially true for startups trying to sell big-ticket items. The average selling price for Aveksa's enterprise access software is $300,000, with a sales cycle of six to nine months. "This is a strategic sell to a head of security," says CEO Taneja.

Saeed Amidi, president of PlugandPlayTechCenter .com, a Silicon Valley tech incubator, estimates it can cost an enterprise software startup from $50,000 to $100,000 in sales and marketing costs for each new account it signs. The rule of thumb, he says, is that such companies need $20 million in venture funding to sign up enough customers to sustain themselves.

That's too steep an admission fee for many entrepreneurs, who are tossing the old sales model. Spiceworks, for example, is having success distributing ad-supported help desk software--no licensing fees involved--while Web 2.0 startups such as Ning are trying to appeal to a mass audience. Andreessen's advice to entrepreneurs: "You're much better off starting a company aimed at a market that wants to adopt new technology, such as consumers, or a self-service model for workgroups or small businesses."

Startups like 1-year-old Techrigy in Rochester, N.Y., start with a few tire kickers, then build from there. In October, Techrigy released its first product, software that monitors what employees and outsiders are saying about your company online. It has its foot in the door of two companies--a consulting firm in the United States and a telecommunications company in Italy--but both installations are modest in size. "The first customer is always the hardest," says Techrigy founder Aaron Newman.

IT departments need to pay attention to the fresh ideas coming from companies such as Aveksa, Right90, Techrigy, and Zenoss. "If CIOs don't take a chance on new technologies, they run the risk of missing out on the breakthroughs that can help propel their business," warns Credit Suisse software analyst Jason Maynard.

Worse, if too many IT departments slam the door on startups, it becomes a self-fulfilling prophesy where innovation dries up for everyone. So do your homework, then give a startup a chance.

>BackTrack < 

---