The risks and rewards of using startups

Several years ago, Rob Duchscher took a chance by signing a contract with a startup vendor of test automation software. Duchscher, now the CIO of Starkey Industries, the largest hearing-aid manufacturer in the United States, did his due diligence. The financials were in order, the company's leaders had a track record of success, and the startup's technology blew the competition away.

A few years in, the company was shut down. And Duchscher was left to pick up the pieces.

"At that point," says Duchscher, "I was hosed."

It wasn't the first time Duchscher got burned placing a bet on a young vendor. Some were bought. Others went bankrupt.

And it probably won't be the last. Unlike some IT leaders, Duchscher doesn't like to play it safe with vendor selection. "A lot of my peers are too risk-averse. They want to make the absolute safest choice," says Duchscher, who was vice president of software engineering and a research and development program manager before taking over IT. "Safe choices lead to a culture of status quo. And status quo, especially today, can make it hard to survive and remain profitable."

Increasingly, IT leaders, frustrated with traditional suppliers, are seeing the appeal of startup partnerships. "The desire for lower risk pushed a lot of CIOs toward the bigger players with deeper pockets and more money for R&D. But they gave up one risk for another," says Christine Ferrusi Ross, research director at Forrester. "They traded the risk that a smaller, newer vendor would somehow fail for the risk that they would be held hostage to a supplier with whom they have no leverage."

It's not just increased influence that draws IT leaders to nascent suppliers. Startups are more flexible with pricing terms and product features. Their contracts take weeks to negotiate rather than months; implementations take months rather than years. Most importantly, they are answering the technology questions that older vendors won't even ask.

"Small, startup companies are the primary drivers of innovation within the IT industry," says Mark Settle, CIO of $2.2 billion BMC Software. "A lot of startups are not only solving the newest problems out there," adds Ben Haines, CIO of Pabst Brewing, "they're solving them at a faster rate."

But working with startup vendors is complicated. They lack the processes and customer support frameworks of their more-established counterparts. Their rollouts and updates require extra oversight. And the CIOs that partner with them have to do a lot of hand-holding throughout the relationship.

"It takes time and direct involvement. But if you're truly getting a unique value proposition that you can't get elsewhere, you have to be willing to invest in this stuff," says Tracey Rothenberger, CIO of Ricoh, the $23 billion maker of copiers, supplies and document-management technology. "We're not [working with startups] for parts of the business that are small or insignificant. We're doing this with vendors that will have the ability to truly be transformative for us." (See "Ten Questions to Ask Before Using a Startup.")

The Upside of Startups

"Startups are a good way to experiment at the edges of your priorities and position your company as an innovator," says William Hsu, co-founder of startup accelerator MuckerLabs.

But a carefully selected startup operating at the core of the business can make an even bigger bang, as Rothenberger found out. "We're trying to get an advantage for our business by moving into progressive technology much faster," he says.

Nine years ago, Rothenberger needed a mobile solution for Ricoh's service force--something that would give it access to parts inventory, schematics and dispatched calls in the field. At the time, the market was ill-defined--many players and even more technologies. Rothenberger teamed up with a new vendor that had a handful of employees. "They had some incredibly good technology, and it gave us a good couple of years with that solution before our competitors had it," Rothenberger explains. "We did it because it was pivotal to our business."

Increased responsiveness and influence over product design are also among the perks of working with a budding vendor. Clorox CIO Ralph Loura works with startups when he's looking for what he calls "systems of engagement"--those tools that take a new, often social approach to a business problem. Young vendors are noticeably more nimble and responsive to his needs as a customer. When Loura asked one startup partner for integration support for a legacy platform, the company got to work immediately. "In an established vendor, it may take 18 to 24 months to make such a request, have it filtered from the field through product management, have it approved for the product road map, and make it into a release," says Loura. "This startup vendor was able to accomplish the same in under four months. [It's] a much flatter organization and the focus is more on being agile than on product lifecycle management."

Duchscher likes the attentiveness. "When you're working with one of the behemoths, you might get a quarterly visit or someone asking you at the end of the fiscal year what they can do for you. That drives me crazy," he says. "With a startup, there's a much tighter interaction. You know the engineers and the owners, and they're actively involved in how things are going."

And, of course, the price is right. If you're an inaugural customer, you may even score freemium rates. "A lot of them don't know how to price their products. One founder asked me if he was charging too little," says Haines of Pabst Brewing. "I had to say, 'Yes. But I like it.' I want it as cheap as possible. If I'm in early, I get that benefit."

Where the Startups Are

CIOs can find viable startup vendors everywhere. Peer networks. Industry forums. Analyst reports. Google searches. "Any CIO that is trying to develop competitive advantage for his or her business through the use of information technology really needs a mechanism for continually scanning the landscape of tech startup companies," says BMC's Settle. Ideally, IT leaders want to get involved "as the [technology] trend is developing, but before the market consolidation happens," says Rothenberger of Ricoh.

One of the best avenues for finding newer, niche players is venture capital firms. Many specialize in a specific type of technology--cloud, mobile, collaboration--so it's relatively easy to identify potential partners.

More important than uncovering the newest player in a market, though, is finding the right one.

Compellent Technologies actually reached out to Duchscher about its new disk array products as he was researching offerings from traditional providers like IBM and EMC. "They were a small shop, and there was concern here at the top that they might not make it," says Duchscher, particularly given that this would be a multimillion-dollar investment and not a sideline experiment. Duchscher did what he always does--get inside the company.

"I like to go onsite and really get to know them," he says. They had a solid, intelligent disk array at a time when others didn't. Funding was robust. The leaders had strong backgrounds. And they were 25% cheaper than other options.

In many ways, running a background check on a new market entrant is similar to traditional vendor due diligence. "The categories are the same," says Forrester's Ross. "But the definitions are different." Financial viability is always important, for example. But if you judged startups on profitability, you'd rule out 90% of them, Ross notes.

"A lot of what we do with established solutions is filter. We have a need. We [find] five or six established solutions in that space and try and eliminate choices," says Loura of Clorox. "This logic needs to be inverted for startups. The question shouldn't be, 'Can I filter them out?' It should be, 'Can I convince myself that they could close the gaps?'"

Those gaps tend to come in three categories: financial, technical and foundational.

CIOs often have certain metrics or guidelines for assessing financial viability. Who is funding them? Are more than half of their users paying customers? How many recurring contracts do they have? How much cash flow is there? "I can't sign up with two cash-strapped guys in a garage," says Rothenberger.

When Haines was looking for a file-sharing tool in his previous role as director of architecture of Red Bull North America, he considered using Dropbox but backed out when he realized there would be no enterprise support. Then he found Box. They had well-known backers. They'd brought in some enterprise heavyweights. He hired them at Red Bull and has since deployed their technology at Pabst.

The technology check is often more rigorous because "there is relatively little third-party validation out there that [a startup] product can deliver," says Nick Gaunt, CIO of the Institute for Innovation and Improvement at the U.K.'s National Health Service (NHS). Rothenberger spends the most time on the technology evaluation. "You may find they have a fantastic concept but it's not mature enough for your organization," he says. Big red flags include a system in perpetual beta or a product road map that shifts with the wind.

Loura looks for portability. "If a solution requires substantial integration or development and is hard to decommission once deployed, then we'd have a greater concern," he says. Solid-state disk storage is less risky, for example, because one could easily migrate back to spinning disks. A new business intelligence and analytics tool is a harder sell. "The investment in end user training, data integration and report building would be significant and not easy to port to a different platform," Loura says. "So we would do more work to become comfortable with the solution."

Then there's the foundation: the business plan, the talent, the leadership. Gaunt of the NHS looks for entrepreneurial veterans with a history of commercial success. "We were thinking about getting involved with a company where the chief technologist was incredibly bright. But he wasn't a good business person," says Rothenberger, who ultimately chose not to work with the company. "They had the superior product, but he had a stranglehold on all the decisions."

Once the Ink Is Dry

One of the biggest issues for IT leaders who bring in these niche suppliers is integration. "It's tough," admits Haines. "You have to make sure your staff is trained in all the different APIs and platforms out there."

Vendor management headaches also increase. "Most aren't very good at managing their vendors anyway, so governance becomes a huge problem," says Forrester's Ross. "You need to parcel the responsibility out to the person who works closely with the niche supplier but also make sure everyone knows how to do vendor management."

Customer service can often be an issue. "Startups don't have a lot of processes in place, and they haven't worked with 1,000 customers," says Hsu of MuckerLabs, who founded BuildPoint, a construction bidding system, at the age of 23. "Support and issue resolution from post-sale to implementation is something you have to work out before the contracts are signed."

Startup employees are often playing multiple roles: product development, project implementation, customer support. "On the plus side, the startup organization is far less bureaucratic," says Settle of BMC. "On the negative side, those same technical resources may already be dedicated to support another customer or to get the next product release out the door, and access to those individuals may be walled off altogether."

If you're a Fortune 500 company used to working with big vendors, you're accustomed to strong vendor account management. "Smaller players have focused so much on their product that that is something they haven't developed yet," Ross says. "The relationship management pieces aren't there." The customer has to take the reins and set expectations for communication.

While the speed of startup rollouts and updates can be a boon, they also require more technical oversight. "We allocate more testing time to the releases," says Rothenberger. "Smaller vendors have fewer controls, so more issues pop up."

There can also be internal resistance in IT to the startup partnership. "The ultimate paradox in most IT shops is that the staff complains that they spend all their time maintaining legacy technologies and that there's no innovation going on," says Settle of BMC. "But when efforts are made to evaluate or prototype new startup products, everyone complains that they already have too much work to do."

CIOs always talk about vendor engagements as partnerships, but that's often just lip service. When working with a startup, parternships are a requirement.

"You can't just make the purchase and go away. It's not like working with Microsoft or Oracle, where you ultimately get a product that works, with a few bugs here and there," says Duchscher. With startups, "things are fluid, decisions are made quickly, and you need to stay on top of that."

At Pabst, Haines is working with some very early stage players--a data integration company that hasn't even entered Round A funding and an enterprise performance management provider going into general release sometime this fall. "It becomes a mini external R&D group for us," Haines explains. "They're giving us a lot, and we're there on the ground level helping them to shape the products."

Ricoh's Rothenberger has even talked to venture capital firms on behalf of three of his startup partners. "It's another more intimate way to get engaged. We get into discussions about where they're going for the next rounds of funding, their last six months of financials, strategies to take the company public or sell," he says. "Initially, you're just buying into the technology. But as the relationship evolves, you want to see the company succeed."

But too much guidance from--or customization for--a specific customer can handicap a startup. "My role is not to push them, but to act as mentor," Rothenberger says. "I'm hiring these guys because their product is unique. If I knew how to build it better than them, I would have done it myself."

The End Game

Startups aren't startups forever.

Loura at Clorox knows a CIO whose entire enterprise is now being run by a major ERP player even though he didn't buy a single piece of software directly from them. The big ERP vendor had acquired the smaller companies he was working with.

It happens. A lot. Loura has seen collaboration tools, social listening platforms, systems monitoring solutions, and more snapped up by legacy vendors or competitors. "Sometimes it's business as usual, but often the acquirer will make changes to the prior road map or strategy," he says.

At Pabst, Haines is waiting to see how Microsoft's acquisition of Yammer works out for him in the long run. "I'm a little nervous," Haines admits. "I've talked to them, and they've been open about Yammer remaining separate, which is good."

Contingency plans are key, starting with securing source code escrow and forensic toolkits in the contract. Even the big vendors can discontinue support for their own products. "It's just as easy for an IBM to make a SKU disappear," says Haines.

At Starkey, that source code and forensic toolkit enabled Duchscher to continue to use the test automation software that was shut down--for a time. "But the life of that product had ended. No more fixes, no more enhancements," Duchscher says. "You can live that way for a while, but eventually if the vendor is not moving their product forward, that impacts your ability to move your product forward."

Today, Duchscher is dealing with the aftermath of Compellent's sale to Dell. "We're not seeing the level of responsiveness that we saw when it was just Compellent," he says. "I'm happy for the people of Compellent, but it was not the best move for Starkey. When I had a real problem before, I could pick up the phone and call the president for immediate action. I don't have Michael Dell's number."

There's also some risk in startup success. "These startups can become big players themselves--not necessarily becoming behemoths, but really owning their space," says Ross of Forrester. "And the vendors can start to take on the characteristics [that CIOs] were trying to avoid in the first place."

Then there are the startup failures. "I've been involved in a couple, and they're always tragic," says Rothenberger. "It's never about the technical aspects of the product. Those things can be fixed. It always comes down to the human element--individuals who make poor decisions or don't communicate well and end up creating a big mess."

Working with startup vendors isn't for everyone. Joe Fuller once started his own a point-of-sale software company for specialty retailers. His employer, Dominion Enterprises, has invested in a startup accelerator program called Hatch. And Fuller has served as judge for Start Norfolk, startup event and competition in Virginia. Yet as CIO of Dominion Enterprises, he doesn't purchase any products or services from startups.

"We are typically dealing with large vendors for our expensive services and products such as disk storage and Internet connectivity," Fuller explains. "These are capital-intensive businesses that small startups really can't compete with."

Fuller wouldn't rule it out entirely. But, he says, "Most startups...are under-capitalized and they can't survive any setbacks. As CIO, I can't afford to go with an innovative company that offers a neat product at a great price if I'm afraid they will go out of business."