Factory Joe

The Fall of Vidoop

When I left Flock in 2006, I blogged the occasion, having helped start the company by contributing a vision for what I thought the web needed: a social browser.

When I was laid off from Vidoop last month, I didn’t so much as tweet about it. The circumstances were different this time. But because the lack of information coming from the company is disappointing (if not frankly irresponsible) it seemed time that I wrote down my recollection of what went down.

Joining Vidoop

I joined Vidoop just over a year ago, in May of 2008.

To be honest, my initial impressions of the company weren’t exactly positive, having first seen Luke Sontag (Vidoop’s co-founder and president, or ” Chief Koolaid Officer ” as he called himself) launch the company with a slick pitch at the Web 2.0 Expo in San Francisco in April of 2007, proclaiming that Vidoop was going to do a rev-share with OpenID relying parties as well as regular users of the MyVidoop service.

I remember muttering to Scott Kveton — friend and chairman of the OpenID Foundation: “Great, someone’s attempting to exploit OpenID already.”

I didn’t think about Vidoop again until February 2008, when I found out that Kveton had joined the company after leaving Strands, a former client of Citizen Agency.

Kveton urged me to reconsider my impressions, suggesting that I fly to Portland in early April and have Luke pitch me on joining the team. Over drinks at Clyde Common, Luke laid out his grand vision for the company and invited me to come work full time on the Diso Project under Vidoop’s stable. Considering the difficult changes I was going through professionally and personally, it seemed like a great opportunity to throw myself into my work without having to worry about finding new clients for a while.

So, with Kveton’s encouragement, I accepted their offer, pulling in Will Norris and announcing the news a month later — news that was met rather warmly.

I would work remotely from San Francisco with Will — our work on the Diso Project helping to raise the visibility of Vidoop in the open source and identity arenas — while Vidoop’s efforts would increase the relevance of strong security in consumer applications built on a raft of open standards, like OpenID.

The Oregon Trail

A month after I joined, I took a trip down to Tulsa to meet the crew for a company all-hands and get a taste of life in the wild wild west (as Luke called it).

I was surprised at the size of the company (around 40 at the time) and the feeling of familiarity ( in the “familial” sense ) among the employees — clearly these folks cared a great deal about each other and what they were doing and were eager to prove to the world that Oklahoma could product high-tech stars too.

The contrast between my initial impressions of Luke’s slick stage presence with the down-to-earth candor of the developers and rest of the company left me positively charged and ready to contribute. There was a lot of heart in the company; perhaps I had judged too hastily?

Joel Norvell — the unassuming CEO and one-time ballet dancer and former chairman and CFO of a jewelry manufacturer — surprised everyone, standing atop a stump to declare that the company would be relocating to Portland. The company would offer assistance for the move, taking care of various expenses and working with families to smooth out the transition. I was impressed by the amount of help offered to the employees — a level of hands-on support that I’d not seen at any company that I’d worked with. It gave me pause, considering the magnitude of asking some 40-odd folks to uproot their families and move across the country to follow a dream: “Would this actually work?”

As it turned out, this was not the first time that moving the company had been proposed. In fact, Luke had enthusiastically suggested a number of other possible destinations, only to be turned down by his more even-keeled CEO. This time, however, with Kveton on board as the resident “Portvangelist“, Luke succeeded.

The plan was to spend the summer finding housing and making arrangements and then leave in September, following the route of the original Oregon Trail in a train of rented campers and U-Hauls, recording, blogging, tweeting and podcasting the hell out of the whole ordeal (Silicon Florist coverage 1, 2 and 3 ) — culminating with a parade down mainstreet in Portland and a greeting from the mayor.

Though thankfully no one died of dysentery (in spite of a few “unforeseen detours”), it turned out that the parade and mayoral greeting were just the beginning of several oversold and underdelivered promises from management.

Shiny, shiny things

Turned out that uprooting and transplanting an entire company from the middle of the country to the west coast is more costly than one might otherwise think.

The decision to switching from PHP to Python also had its costs, primarily on developer focus (even if the decision played to the strengths of the dev team). Between the upheaval of the move and switching development modes — the team lost a lot of time, resulting in missed deadlines, unfinished projects and lost contracts.

Merely two months after cozying in above Backspace in Portland’s Old Town, Vidoop announced “changes” — a euphemism for layoffs.

Blaming the economy, Vidoop shed 11 employees. Blogging as CEO, Joel prudently pledged to “concentrate on those areas that make the most sense in this economy.”

As part of the changes, they also cut me back to half-time — leaving me with rent money, health insurance and a shaky future with the company.

With a new emphasis on product, my work on Diso was scaled back and I was tasked with redesigning the company website, working on the UI for VidoopSecure and helping to architect the user experience for the-silently-launched VidoopConnect product (now defunct).

I traveled to Portland several times during the winter and spring. We were making considerable progress with a renewed sense of urgency and will to execute. The VidoopConnect team in particular, lead by Adam Lowry and Michael Richardson, was firing on all cylinders.

Some time in February, however, the focus again began to blur.

On top of the other half-finished products in Vidoop’s quiver, VidoopCAPTCHA was introduced a lightweight mechanism to promote the ImageShield — as well as attempt to dethrone ReCAPTCHA. Development on VidoopConnect ceased. Will Norris was reallocated to VidoopCAPTCHA and implemented a plugin for WordPress while the rest of the team produced an API and web service.

The shift from VidoopConnect was abrupt and unprecedented: just one more example of the chaos of Vidoop’s product strategy.

Now, I haven’t yet mentioned MyVidoop, Vidoop’s much-loved consumer OpenID provider and browser-based password manager. That’s largely because it too was deprioritized around the time of the first layoffs — kept alive on little more than life support. In some ways, the fate of myVidoop is both emblematic of the lack of focus and thorough execution that I believe contributed greatly to the fall of Vidoop, and remains one of the more problematic pieces of this story (more later).

The beginning of the end

This past March, I paid my own way to SXSW. Meanwhile, Vidoop picked up travel for Kveton (by now some kind of VP of Engineering Open Technologies), Sontag, Matt Selbie (VP of Marketing), and Scott Blomquist (CTO) who all shacked up in some sweet pad somewhere outside downtown Austin.

At BarCampAustin, Kveton presented on Bacn.com, a scrappy side-project he’d launched with Richardson that — get this — sells real bacon — and that took off with a life of its own. Turned out Vidoop didn’t take too kindly to Kveton telling the story of Bacn when he apparently should have been pimping their tech.

Upon returning to Portland, he found out that he’d been fired.

This was — for me as I’m sure for others — the beginning of the end.

Since Kveton had been helping to coordinate my work on the Vidoop side of things, I lost my representative at the company. Though I kept in close contact with a number of the developers, I ended up busying myself with speaking, traveling, and evangelizing OpenID on the road.

And then came the rumblings of financial problems. On April 15, I received word that the company was about to close a major round of funding — and just in time, too, because Vidoop had failed to make payroll.

Two days later — April 17 — I was in Sebastopol at Social Web FOO Camp when I received a phone call from Wiley Parsons, Vidoop’s CTO, with instructions to cut up my company credit card.

My first thought was that Vidoop had been sold, an idea later dismissed. To this day, I’m still not entirely sure what happened, but it was obvious that things were being kept together with chewing gum and shoelaces, threatening to split apart at any moment.

The final straw

On May 11, I was in Paris in the middle of a trip bookended by speaking engagements in Hamburg and Belgium. Luke messaged me, requesting that I call him ASAP. When I did, it turned out that Vidoop was out of money, unable to make payroll again. All but a skeleton crew remained. Luke assured me that he was going to keep fighting — attempting to sign some final deals to save the company. He promised to share more information by the end of the week but the only public statement that’s been made to date was a mealy-mouth blog post with no discernibly useful information whatsoever.

Two weeks later, Joel sent out what appeared to be heartfelt last rites for the company, again lacking any actionable information. A day later he clarified things, posting to Vidoop alumni mailing list that, “Vidoop LLC is officially out of business. Unfortunately, there are no funds to pay the unpaid wages or other liabilities.”

This was the memo that leaked to TechCrunch.

During the last two weeks of May, there were two parallel tracks of activity — segmented into those left fighting for whatever was left of the company — and those who’d been laid off. The unemployedstruggled to find new work, attempted to secure new Visas, grappled with the arcane unemployment and healthcare systems, and tried to pry more information from the company; others called it quits and moved back to Tulsa.

Those left fighting for the company — Luke, Blomquist, and Selbie — have provided strange accounts of what they’re doing to try to rescue the firm or — apparently — start something new.

There’s been no public acknowledgement of the situation, either on the company blog or on Twitter, save for this melodramatic post (presumably from Luke): “dead, no. bloody, yes. still got fight left. details soon…”

The media and wider community have been left to their own devices (and Arrington’s original post) to fill in the blanks. With no information forthcoming from the company, it’s impossible to know what to plan for, or what the fate of services like MyVidoop will be.

OpenID, myVidoop and unfulfilled promises

I’m writing this post not because I’m bitter — most startups fail and I knew this when I joined the company — but because the lack of information coming from Vidoop has been irresponsibly minimal — both publicly and privately. Vidoop has failed to speak clearly and consistently to the community that it pledged to serve and secure, and it has failed to produce definitive information for the staff that risked much and poured themselves into the company.

The failure of Vidoop was a failure of business and focus, not technology. It’s as simple as that. And yet, with the company out of business, those responsible for communicating clearly and transparently about the health of the company are failing to do so.

This is bad for several reasons. Among them:

  1. First, those who continue to use MyVidoop have not been told what’s going on or planned for the service, and can therefore not make an informed decision about whether they should continue using the service. If they do — with the presumption that Vidoop is solvent — they may well end up locked out of not only their MyVidoop accounts, but also the accounts that MyVidoop holds the keys to.

    None of MyVidoop’s Terms and Conditions mention the availability of backups or access to data should the service be shut down — nor does it state what will happen to the servers that house personal data should the company file for bankruptcy. These are questions that Vidoop needs to have answers for, along with advice for people that want to leave the service. In contrast, Yahoo has taken aggressive steps to notify people a full 60 days in advance of the pending closure of the Yahoo! 360 service, providing guidance and tools to export user data.

    I discussed the fate of MyVidoop with Scott Blomquist when I was in Portland for WebVisions. I implored him to issue some kind of public statement about the welfare of the MyVidoop service, but so far he has demurred. He has proposed running the service as a “community project” and has attempted to enlist the help of several ex-Vidoopers, but I’m not convinced that his plan will hold water. Besides the trademark issues wrapped up in continuing to use MyVidoop.com as an OpenID provider, there are various licensing fees associated with keeping the strong authentication services functioning (voice confirmation, SMS-verification); and, even if those services were turned off, running a high-security service as a community initiative just doesn’t add up — not after all the fear-mongering that Vidoop does on its Twitter account!

    The responsible thing to do would be to announce the closure of the MyVidoop service in 30 days (if the servers can be kept running and secured for that amount of time); the “community project” idea should be scrapped in favor of building tools to export people’s data, and if it’s amenable to a company like Janrain, the MyVidoop users could be invited to move their accounts to MyOpenID.com (presuming that Janrain has reasonable data back up, export retention policies!) Otherwise, keeping the service running longer will just invite catastrophe with no one on the payroll to deal with issues related to security or uptime.

  2. Second, this whole situation paints a negative caricature of a member of the OpenID community. It’s not the technology that isn’t sound or market-worthy — it’s that perception is oftentimes nine-tenths of the law. Vidoop’s drawn out failure highlights the risk and importance of choosing a responsible identity provider — or of using delegation — the feature of OpenID that lets you use your own domain as your OpenID, but assigns someone else the responsibility of serving your identity. Vidoop’s failure could be seen as a blight on the technology and community if people fail to recognize that the problem with Vidoop was not OpenID, but was the same thing that’s caused countless other companies to fail: the inability to focus and execute in a consistent and thorough manner.

To its credit, in its time, Vidoop assembled a cadre of highly talented and motivated folks who really did want to make their mark on the world, the right way.

Projects like Identity in the Browser (“IDIB”), Emailtoid and EAUT, VidoopConnect, and VidoopSecure exemplied solid and well-engineered solutions and proofs of concept that demonstrated an open, innovative approach to delivering secure solutions to the consumer web. To have those efforts sullied by the incongruous response from management is unfortunate and unfair.

With so much promise, it seems sad that Vidoop could fall so far, so fast.

Update: Scott Blomquist has posted to the Vidoop blog, stating:

Despite the claims of others elsewhere on the internet, myVidoop will continue to run for the forseeable future. Those of us who are still at Vidoop are committing to give you at least 30 days’ warning in the event that shutting down myVidoop becomes necessary.