This is part three of ProgrammableWeb's series on the End of the API Economy as We Know It. In part two we offered a case study on how PeopleLinx was able to thrive even after LinkedIn cut off access to its APIs.
Taking advantage of the power that APIs offer is a quick and inexpensive technique for a business, provided the APIs it depends on are reliable. But if any of those APIs change or go away, havoc can ensue. As ProgrammableWeb often reports, such occurrences are not rare. Changes to an API happen all the time — just take a look at the change logs for Facebook, LinkedIn and YouTube and other popular APIs.
In a previous article, "It's the End of the API Economy As We Know It", we provided an overview of the problem. Now it's time to tell a story of Stitch, a company that found a way to succeed, despite the uncertainties inherent to the API economy.
From the Attic to the Boardroom
Jake Stein had about $40K in the bank when he and Bob Moore left well-paying jobs at a New York City-based Investment firm to start their own data analytics company, RJMetrics (the company Stitch eventually will spin off from). Stein knew he'd need to use his savings to pay his modest living expenses while their startup got off the ground. Building a customer base takes time, and generating the revenue to pay Stein and Moore a salary would take even more. Stein lived on his savings for 18 months until he could draw a paycheck, but he strongly believed in their work. By the time Stein and Moore began drawing a salary, RJMetrics was on its way to a run that lasted years, beating the very high percentage of startup failure (60% to 80%, depending on your choice of source).
Stein and Moore started RJMetrics in 2008 from an attic within Moore's house near Philadelphia. Eight years later when the company was sold to e-commerce Platform provider Magento, RJMetrics had 100 employees on its payroll along with the attention and resources of investors such as Jason Goldberg, Uncork Capital, Lear Ventures, Trinity Ventures, and August Capital.
Stein and Moore's success was fueled by their new take on data analysis across sources. They had great success selling access to their analyses to companies eager to fork over the money for that intelligence. Stein, Moore, and their feisty startup, RJMetrics, were early beneficiaries of the API economy.
Their story is compelling not because they were victims of the volatility and risk when relying on third-party APIs as a business's mainstay, but rather because they figured out how to adapt to the uncertainty and prosper.
"I want to be close to this. It's super fascinating to me."
Stein liked tech, but he was not a real geek. The two years of C++ classes he took back in 1999 while attending Montville Township High School still interested him. The classes gave him insights into the details of programming necessary to move from tech user to tech maker. Also, he liked the experience of control over the machine that programming provided, but, he wasn't a standout like one of those kids destined for a Computer Science degree from MIT or Stanford. Yet, while other boys his age at the suburban New Jersey secondary school had their noses in magazines such as "Sports Illustrated," "Rolling Stone," or "Car and Driver," Stein was reading "Forbes." By the time he got to the C++ classes he'd also become a regular reader of Slashdot. He liked pure tech well enough, but his real interest was more about the business of tech. "I never thought, omigosh, this [tech] was what I want to do with my life. But, [instead I thought] I want to be close to this. It's super fascinating to me," he says.
Stein graduated high school in 2002 and then did his undergraduate work at the University of Pennsylvania's Wharton School, majoring in Finance and Entrepreneurship. He graduated from Wharton in 2006. His goal at the time was more about pursuing his interest in the intersection of business and technology rather than making a lot of money. "I always just assumed I'd be fine. I don't have particularly expensive tastes. [Even today] I live in a one-bedroom apartment today with my girlfriend. I had the luxury of not worrying about money because the thing I was interested in the most, the intersection of business and technology, is something where I knew I could make a fine living," he says.
Stein's first job out of college was with Insight Venture Partners (IVP), a venture capitalist firm he worked for the previous summer as an intern. So, getting a fulltime job was a lock. Stein showed up for work at the investment firm in September of 2006. It turns out the gig at would be not only his first traditional "job," but also his last. He was to never again work for a company that he did not run and in which he did not have an equity position.
IVP's mandate was to invest in software and internet companies that showed potential, which meant everything from infrastructure software to applications to Threadless.com, a company that sells funny T-Shirts. Stein started as an entry-level analyst and spent his days looking for investment opportunities, and once he found one, diligently managing the changes that company needed to make before receiving IVP's investment.
Part of Stein's discovery process required him to perform an extremely thorough data analysis on companies of interest. IVP's focus went beyond reviewing tax returns and banking records. The company used computational data analysis to determine metrics such as customer concentration, customer lifetime value, customer retention, and overall ROI.
IVP gave Stein a practical, real-world education well beyond the theories he studied in college. "I took a marketing class in college. It was very much about branding and feelings and stuff like that. And then I found out there was this whole domain of Google AdWords and SEO and targeted lead-gen and webinars and all this stuff," he said.
During his tenure he learned how to manage a sales team comprised of Sales Development Representatives (SDRs) and account executives. They introduced him to, the tiniest details. "There's these ratios about quotas to how much they actually bring in, and bookings and billing. All of this data could be PUT into a spreadsheet [and analyzed]. But I had only known, 'You want to be a good salesman? Then you want people to like you.' or 'You want to be a good marketer? Then you want to have a really catchy tag line.'
Stein found that in the real world, these companies are much more technical and systematic about sales and marketing than he'd been taught in college. "It was really fascinating to me," he says.
In addition to soaking up the latest knowledge and technology used in marketing and sales, Stein was making important contributions to IVP and, as such, was making good money, but as he encountered the ups and downs that were part of the job he realized that for him the frustrations of the downs were greater than the benefits of the ups.
"I just don't know if I want their job."
Around mid-2008, Stein was putting together a number of investment deals for IVP, one of which he found particularly exciting. The company under review was an excellent fit with IVP's investment thesis, and he had a great rapport with the founder. Great success for all parties was on the horizon. Then it all went south. Within a week that deal and every other deal Stein had been working on fell through. Despite this demoralizing experience, Stein felt confident that he could put it all back together again, but then he realized he was sure he wanted to invest the time and effort this rebuild would require.
When he looked around the company, he saw people (many who were closes friends) who'd been working there for years, even decades, grinding away at an endless cycle of deal making, deal breaking, and damage control. Stein respected the work they did tremendously, but when he considered his needs, he had to face facts. "I just don't know if I want their job.", said Stein. Shortly after his realization, Stein met a kindred soul, Bob Moore. Stein had been kicking around ideas for a business, but Moore was a bit ahead of him and already had found his niche — he wanted to automate the process of due diligence.
Doing the Mundane but Necessary
For IVP and many other companies in the venture capital space, due diligence was a labor intensive, mundane activity. Stein and Moore knew this from direct experience. Both had to suffer the torture of acquiring data and then making sense out it. Data came from all over: text files on an FTP server, encrypted DVD backup sent via Fedex, or Excel files emailed to them directly. Sometimes the analysts were even given direct access to a company's database. Once they had the data, then they crunched the numbers, which could take days. Also, Stein and Moore had no guarantee that they had the latest data. More than once, a company under review called up and asked them to rerun the data analysis process to accommodate new data. The entire process was brittle, and it was time-consuming.
Moore believed they could simplify the entire process by applying standard data analysis techniques. Moore reports, "It didn't necessarily seem obvious the first time I worked on a due diligence project. But by the fifth time it became clear that 80% of the work was all the scaffolding around the analysis and not the analysis itself. And that was when the insight came to automate the vast majority of that process and turn it into a product."
By March of 2008 Bob started to socialize the idea with the younger employees at IVP to create a software application that replaced the mundane parts of the job. Stein was one of the employees who listened to the pitch. A few days later he went over to Moore's desk and said, "Hey Bob, this is a pretty good idea for you to quit [IVP] and go do this. Wouldn't it even be a better idea for both of us to quit and go do this?"
So they did.
Realizing an Opportunity
Stein left IVP in September of 2008 to start RJMetrics with Moore. (He remembers the time clearly because it was during the week that Lehman Brothers went bankrupt.) Each contributed half of the $15,000 of working capital they projected they'd need to meet expenses while they grew the company. Neither intended to draw salary.
They made sure to share the workload. Both were well versed in the business. Moore was more advanced in terms of programming the advanced features of the application, but Stein had enough coding experience necessary to make contributions. As Stein reports, "I built the website. I did a small percentage of the UI, which is basically taking some open source stuff and customizing it. I pulled on open source RSS reader, took out almost everything and that became the front end of our dashboard. But [Moore] was really the driving force bend it all," Stein said.
By January 2009, five months after they founded RJMetrics, they received their first client payment — something like $200, according to Stein. By the end of 2009 revenue had grown to around $50,000. The following year sales grew to about $200,000. "We had some really nice growth. In the scheme of business it was relatively small, but it was growing fast," he said.