A startup that’s operating in stealth mode raises an almost stunning amount of money before it releases a product. Investors write outsized checks to the outfit anyway because of the people involved in it, but before you know it, poof, the company is imploding, and the capital is gone.
It’s a story that industry watchers know well at this point. Clinkle — the payments reward network that raised what was at the time the “largest seed round in Silicon Valley history,” then never released a product, remains the most widely mocked example of the genre. Another related flop is Airtime, a video chat service that was created by Napster founders Sean Parker and Shawn Fanning and launched with so much fanfare that you just knew it wasn’t going far.
The latest example to follow a similar trajectory: Primary Data, a four-year-old, Los Gatos, Calif.-based data virtualization startup, is in the process of shutting down after raising a whopping $100 million in equity and debt and attracting the likes of Apple co-founder Steve Wozniak to its management team.
What happened? Neither the company nor the numerous investors to whom we reached out over the weekend have responded to our requests for comment. But according to a trusted source close to the company, Primary Data’s problem from the outset was that its technology was never quite as compelling as it needed to be, given that it was trying to sell mission-critical software. (If it’s not up to snuff, data virtualization software can create challenges with manageability, usability, data quality and performance.)
Certainly, the issue wasn’t one that Primary Data’s investors — including Accel Partners, Battery Ventures, Lightspeed Ventures Partners and Pelion Venture Partners — anticipated at the outset.
With Primary Data, VCs were re-investing in a team that had brought them financial success with their previous startup: Fusion.io, a flash storage company that enjoyed a highly successful public offering in 2011. Indeed, when Fusion.io co-founders David Flynn and Rick White left in 2013, investors quickly provided the pair with $50 million to spin up their next thing. (It was separately becoming clear that Fusion.io was heavily reliant on two big customers. As those customers dialed back on their flash storage, Fusion.io’s share price began slipping, and it was acquired in 2014 by the chipmaker SanDisk.)
We aren’t sure as of this writing why Primary Data’s software disappointed. What we do know is the company had brought aboard the founders’ longtime colleague, Lance Smith, as Primary Data’s new CEO in 2014. Smith had joined Fusion.io as its president and COO in July 2008 and spent six years with the company.
Immediately upon joining Primary Data, Smith realized that its burn rate was out of control, particularly for a company with no revenue. But while the processes Smith instituted helped, they didn’t change the fact that Fortune 500 companies weren’t prepared to buy Primary Data’s technology — even after Wozniak joined the team shortly afterward as chief scientist. (Like Smith, Wozniak was also an alum of Fusion.io, where he’d been named chief scientist in 2009.)
We’re told that the offerings from an Israel-based storage technology called Tonian, which Primary Data acquired for undisclosed terms, were expected to help address the problem but fell short. Meanwhile, a new version of Primary Data’s software that was rolled out in August may not have been sufficiently riveting to bring aboard key customers, either.
In fairness, Primary Data and other data virtualization upstarts face an increasingly daunting uphill battle, as virtualization giants like Dell’s VMware have continued to grow stronger and the data storage market more broadly continues to consolidate. (Even Cisco recently sold its data virtualization business for undisclosed terms, saying it was “not aligned” with Cisco’s long-term focus.)
Still, Primary Data’s lofty valuation out of the gate also apparently worked against the company. Though it announced funding this summer from insiders — it announced $20 million in follow-on funding and a $20 million line of credit that we understand was basically a bridge loan provided by its founders — we’re told that its backers more recently decided they’d rather shut down the company than re-invest on terms they found disagreeable.
Specifically, they were asked to allow their preferred shares to be converted to common — and then reverse split 20:1. (In other words, management wanted to reduce the total number of shares outstanding and increase the share price by that same multiple.)
The VCs said no. Soon afterward, the company’s website went blank.
Reached last night via email, Wozniak said he hasn’t been “on top” of Primary Data’s day-to-day goings-on, largely owing to his public-speaking obligations.
He also said that he proudly maintains his status as the company’s chief scientist, while acknowledging that he’s “not in the loop about such things as re-financing or shutting down.”
Wozniak, who is frequently described as one of the nicest people in tech, further added that he does consider himself “a part of Primary Data and would do whatever I could, even without salary, to help them.”
If “PD shuts down,” he told us, “it’s very sad because of how much good technology has been developed by them for so long.”
Pictured above: Co-founder David Flynn pictured on Primary Data’s site in better days (November 2017).