Code Stars: How a small group of young engineers is pushing Edmonton’s startup scene forward
“If you want to make sure [you have traction] and a good, sustainable business, build it out of Edmonton”
by Tim Querengesser
while(1) do x = 4*a*x*(1-x); p x; end
About four months ago, Jason Smith started seeing flaws in the machine. “No company should fall apart because of too much business,” he says. The software being developed by the startup he worked for promised to fatten margins for online retailers, and in theory it delivered. But Smith saw that the software, known as “Version 1” and developed by outsourced engineers, contained a flaw: it needed too much back-end work to be efficient. The software gave the company a product to sell to drum up interest; the company, in turn, hoped sales would bring a financial injection, which the company would use to fix the software. But while the money did not come and the fixes were not made, the sales pitches continued. Smith realized that if usage spiked, labour requirements in Version 1 would be too high.
In the weeks that followed, Smith, who asked to remain anonymous for fear of repercussions, says developers left the startup and marketers started replacing them. Version 2 was written by other local engineers, to fix Version 1. But, for various reasons, the company halted development. Smith knew Version 2 was the only product that could deliver, let alone make money – and it was frozen.
So, three months ago, Smith parted ways with the company. But he wasn’t too concerned, since he was confident something better was around the corner. And that something is not what it might have been in years past – not a secure, high-paying corporate IT job. Instead, Smith is convinced it is in another Edmonton tech startup. His confidence comes partly from who he knows. Smith is in his early 30s and part of a community of Edmonton coders who have become bored with corporate IT work. The group numbers in the dozens and that small size “tends to make giants or rock stars out of people who stand out,” Smith says. The community is young and engaged – the sort of people who program for fun as well as work. They have skills desperately needed by IT departments; that means they can selectively pick where they move. And where they land when they move is influential for the success of tech companies across Edmonton, especially startups.
Mike Deering moved back to Edmonton from Vancouver in 2005, but kept working remotely as a technical contractor for international companies until 2012. “Within the last year or two there’s been some real opportunity and real movement among those [Edmonton] people who were known globally,” says Deering, a 36-year-old coder who now works with Edmonton’s Mitre Media, a financial information tech startup. After he moved back, to get married, Deering says he noticed the city’s tech startup scene was blooming. So, rather than return to Vancouver or continue working remotely for international companies, in 2012 Deering decided to join his own city’s startup scene. And he says other coders like him, with international clout, are being pulled into it, too.
This pulling is essentially a process of talent clustering. Ben Zittlau, a 28-year-old programmer with Mover, one of the city’s current startup stars, says prominent members of the coding community tend to hire from within when the companies they work for expand. “What you end up with is a consolidation of this community in a handful of companies as we recruit developers to join us,” Zittlau says. For tech startups, however, the group’s tendency to recruit from within presents a double-edged sword. If a startup can become an attractive “home” for the coding community, the resulting talent surge it can rely on “likely accelerates the progress of those startups,” Zittlau says. If it can’t, the startup’s prospects are decidedly less promising.
Smith sees this as something like quality control. The tendency amongst the coders to choose a few winners and many losers is positive for Edmonton’s overall startup scene, he says. He recalls the trajectory for Jobber, formerly OctopusApp. The startup once struggled to attract workers, he says. Then, in 2011, the company’s co-founders, Sam Pillar and Forrest Zeisler, switched their product focus, renamed the company and landed a private capital injection. And in the interim they have attracted six coders from the group and are building a better product, Smith says. “The real learning comes when someone tries again,” he says. “Any time there’s a shift, overall, the local startup community tends to benefit. As a startup [in Edmonton] you have to be exceptional on some basis – technical, cultural, social – and when things start going badly, people jump. The good people leave while the lacklustre ones stay. The value isn’t in leaving a trail of failures behind but in trying to leave as short a trail as we can with as few bodies as possible.”
Back in 2013, a website for an Edmonton tech startup called Zenlike appeared, then quickly disappeared from the Internet. Its co-founder, 33-year old David Quail, says Zenlike’s product, an app, relied on artificial intelligence to automate life’s more boring tasks. “We saw some traction and people loved it,” Quail says. Traction, of course, is a tech product that has customers; the more traction there is, the more real value a tech startup has. Quail says Zenlike had that traction, but its proposed features, which relied on artificial intelligence, were too advanced to grow from an idea to a full-scale company in Edmonton. “The major learning with Zenlike was that people get this, but how the hell are you going to build it? You’re based out of Edmonton.”
Zenlike is not a company that Edmonton’s coding group abandoned. Instead, Quail left. He started talking with SRI, the San Francisco-based tech company that built apps like Siri for the iPhone and is one of the world’s leading R&D institutes. They were intrigued, and he’s now an entrepreneur-in-residence there. But Quail’s partner, his two children and many developer friends remain in Edmonton, so he splits his weeks between the two regions. This offers him a unique perspective on startups and recruiting talent. “You could argue that some of the world’s best engineers are working [in the Valley] at Google, Twitter, blah-blah-blah,” Quail says, “but if you were to look at the mean-average for engineers, from my experience, Edmonton has stronger developers.” The Alberta Innovates Centre for Machine Learning at the University of Alberta is a big factor, he says. “That was one of the reasons we thought Edmonton would be a fantastic place to build Zenlike,” he says.
Edmonton’s problem, Quail says – both during Zenlike’s early stages and following his incubation stint at SRI, after which he hopes to have a company with a product to sell – is the potential to scale its top talent to a larger level. “We’ll eventually spin out a company, if all goes well, and the first question is where it’s going to be based. I would consider Edmonton, but I think that would be a hard battle with SRI. They typically want their companies based out of the Valley.” The problem? “It’s developers,” he says. “There’s a pool of amazing developers [in Edmonton], but I don’t know how deep it is.”
But Quail notes an irony in that situation. The Valley is a vacuum for engineering talent, pulling graduates from everywhere, including Edmonton’s elite, the graduates from U of A’s machine learning cluster. “They’re all going south,” he says. But the Valley, he says, is also a hype machine. Startup stories such as WhatsApp’s acquisition by Facebook for $19 billion have created a gold rush type fervor. Quail says this is leading to an “incestuous bubble of false positives” for startups. Recently, he wrote a blog post pointing out that money, buzz and talent do not mean a startup has actual traction. He argued that there is value in taking an idea out of the bubble and into the real world. “If you want to make sure [you have traction] and a good, sustainable business, build it out of Edmonton,” he says.
Still, while he says he would like to do that with Zenlike, Quail is not sure he would be able to. And he is clear that to succeed here as a startup, a company has to be endorsed by the coding community. “It’s incredibly competitive,” he says. “In Edmonton, it feels like there’s this really strong pocket of engineers, and they all know each other, they’re all connected, and there’s a lot of influence that they can have on each other to move about. If you can position your company to be favourable in their eyes, that’s huge. If you can’t, you’re stuck trying to recruit C- or D-level developers, because no one will touch you.”
Who are Edmonton’s code stars? They are a group, almost exclusively men, many of them freelancers, who gather at regular meetings to discuss development platforms like Ruby on Rails, Go, Django or Python, or methodologies to programming, like “kaban,” “agile” and “lean.” The meetings are a recent creation, formed over the last five years. Mark Bennett is a key founder. Bennett is a freelance programmer who returned from Australia in 2009. He’d been excited by the Ruby on Rails community there, only to find something similar did not exist in Edmonton. So, Bennett and a few others started YEGRB, a meetup that aimed to connect Ruby developers in Edmonton. The YEGRB group has helped spawn several meetup groups for other programming languages or methodologies. And these groups have in turn become the place to hire and find jobs – indeed, their web pages now include job listings. “We now regularly help connect developers and entrepreneurs looking to grow or launch a team,” Bennett says.
For Smith, the meetings are where movements in Edmonton’s talent pool for startups begin to happen. Companies, both startup and corporate, scout for new hires; group members try to convince friends to move from something boring to something challenging, or even start something together. And newbies are now attending, driven mostly by word of mouth, with many saying they are unhappy with their corporate gigs. “The moment someone shows up to a meetup, they’re already on the path,” Smith says. “If they’re looking for work, they’ll be snapped up in almost no time.” For Bennett, the meetup scene is a “virtuous cycle” that sees older members educate younger ones, which helps the city’s talent pool “run at the front of the pack” embracing changing technology. But for Smith, the scene has done more than elevate skills in the city. He says what’s critical is developing the community’s confidence. He notes that today, many Edmonton coders are working remotely for companies in Chicago, Los Angeles and Portland, and that this sees the community, by extension, increasingly less likely to be hoodwinked – either by a sexy startup with an emptying office or their own idea, which they have kept to themselves rather than expose to scrutiny. In the past, “there was probably a lot of [developers] wasting away in their basement because they hadn’t brought their idea to a group for criticism,” he says. Today, the group has what Smith calls a “strong BS meter” for hype without substance.
For Bree Emmerson, the 29-year-old head of the Edmonton Chapter of Ladies Learning Code and a freelance web developer, the city’s coder scene is a critical social hub. “People are starting to recognize each other,” Emmerson says. “We’re not necessarily afraid to ask for help and we realize we can’t do everything ourselves. Even though we’re solitary and might be competitors, we still want each other to succeed.”
Still, all is not perfect. Emmerson moved to Edmonton from Kelowna in 2009, and says one of the reasons was the bigger potential in the city’s tech industry. But when she arrived, she noticed a lack of female coders in the community. So she started the Edmonton chapter of Ladies Learning Code, a national organization with 17 chapters. “I have more than 30 female developers [in Edmonton] in my address book,” Emmerson says. “The problem is that they are less visible. The industry has a problem with criticism, isolation and sometimes harassment when women are in an office filled with men.” She says that while the female coding community is bigger in Edmonton than people assume, “It could always be better – I have 60 male developers in my address book.”
Regardless of the community’s flaws, Edmonton needed its emergence to happen. The city’s first wave of tech startups took root in the mid-1990s, and they came from people who are now like legends within the coder community – Ray Muzyka, co-founder of BioWare, being the most prominent example. After that, youth social-media site Nexopia rose up, along with the financial website Investopedia; both were bought by U.S. firms. Since then, however, there hasn’t been a headline grabbing startup tech success in Edmonton. Yes, Yardstick Software and Jobber have grown from startups to become secure places to work for engaged techies who want to push the boundaries while still collecting a reliable paycheque. But there are only so many jobs in places like these, and to keep opportunities open, lest coders be absorbed into the government cubicle farm, there needs to be more startups hitting explosive growth. Those that get frustrated waiting for them, many Edmonton coders say, often pick up and move to places with more opportunities – Vancouver, Toronto, Waterloo or even Silicon Valley.
But still, there is a shift taking place that is opening spots up for talent to land. Many are noticing. “Half the people we see come to Startup Edmonton now are people who have jobs in [corporate IT] and want to work on something else,” says Ken Bautista, the organization’s co-founder and himself a tech startup entrepreneur. “They come to a hackathon [a Startup Edmonton event where coders show off their ideas] and they’re like ‘Awww, now I have to go back to real life.’ And I’m like, ‘What would it take to make that your actual job?’ ” Bautista, who helped found Startup Edmonton as an incubator in 2011, says bright talent being “locked up” in the high-paying but low-impact world of government IT and the energy sector, has long been Edmonton’s Achilles heel.
And some of those who have left for greener opportunities have returned to Edmonton. Take Deering, who took a job in Vancouver because he did not want to work in an IT department for government or the energy sector. He became a contractor for international companies, and gained a name inside and outside Edmonton. But now he’s back, and his movement is likely convincing others that they should stay, too.
onNewInsight.listen( (learning) => broadcast(learning, to: yeg) )
Daniel Huckstep calls the life and death of a startup’s idea “pivoting.” He is one of the Edmonton code stars, as he leads the regular Go meetup (he’s even writing a book about the language), which gives him a high profile, and he’s a natural programmer, learning as much as he can in his free time. He notes that big companies like eBay, PayPal and others all pivoted from their original configurations as startups before finding success. “It’s sort of a friendly way in the startup world to say ‘Well that didn’t f***ing work. Let’s try something else,’” he says.
But Huckstep says startups often fail to pivot. Among the many companies he has coded for in Edmonton is CodeBaby, an “intelligent virtual assistant startup” formed by the founders of BioWare in 2001. Huckstep says the company still calls itself a startup because it has not grown or found enough customers. In 2011, when the company was struggling with money, Huckstep jumped.
While Huckstep chose Yardstick Software, which grew out of a startup, he says many in the coding group are attracted to good businesses underpinning innovative ideas. And he says the mixture of those willing to take risks with those demanding solid companies is necessary for a startup scene that’s evolving – with startups that are pivoting. But the reason he sees Edmonton’s startup scene gaining strength is down to movement. To keep the good people, you’ve got to find a real business idea that’s going to pay. “It’s probably a good thing,” Huckstep says. “There’s probably less time wasted on stuff that’s not going to work.”
Smith agrees. “Not only is iteration key for a product’s eventual success, but it’s important in also shielding a [startup] company from its own dogma,” he says, reflecting on his own recent movement out of a startup that, from the outside, may look like a potential success but has likely found its growth peak far too late in the game. A startup can often find a product-market fit, Smith explains, but if it does not challenge this fit, it can plateau. “If a company finds any amount of success in an early version of a product they can either assume it to be a perfect fit, as-is, by their divine genius, or they can question how to dig deeper or fit even better with a larger market, and invest back in themselves in search of that next working iteration,” he says. Movements of big people rattle those assumptions. They signal that there needs to be another pivot. “At some point, if a company wants to sell to a higher-paying or larger market, they have to treat how that product gets built with more rigour and seriousness,” Smith says. “Fearing that next risk, which might prove their small success has an upper limit, or trying to get even larger results with the same tools that got early results, will probably not get a company much further.”
SIDEBAR: Where are the students?
Tech startups feed on brains. Waterloo, San Jose, San Francisco and other cities that are technology hotbeds all have schools that are “pumping out students like crazy,” says David Quail. “If Edmonton and technology are really going to take off, U of A has in some ways got to lead,” he says. The university has to a degree, through the Alberta Innovates Centre for Machine Learning (AICML), a world leader in artificial intelligence (AI) research. But Quail says it’s a numbers game. To launch his Zenlike startup in Edmonton, Quail would need dozens of AI people; the trouble is, graduate numbers through AICML are small and falling: in 2011, 61 students graduated with undergraduate AICML specialization. In 2013, only 48 did.
The trend continues in Edmonton’s colleges. Over the last three years, two of NAIT’s technology programs – the computer engineering technology diploma and the bachelor of applied information systems technology – saw decreased enrollment (though two other programs saw increased enrollment, and one other was stable).
Andrew Bryson, president of Quercus Solutions, says self-education is partially filling the gap. People who want to work in technology can teach themselves through online tools. Tellingly, he says the most cutting-edge platforms are being used and discussed online, and are often overlooked on school curricula. But while Bryson says he is happy to hire someone if they have the skills, regardless of education, he prefers school-educated workers. “You’re going to get a little bit better organizational behaviour and communications skills,” he says, “because they’ve taken courses, they’ve practised, they’ve trained. So there’s definitely value to school.”
Correction note: This version of the story has been changed from the print version to clarify the role Jason Smith played in writing Version 2 of the software.