Sam Presti made a risky bet in the summer of 2018. The general manager of the NBA’s Oklahoma City Thunder traded for star player Paul George.
There was just one problem.
George had one year left on his contract before he was free to sign with whomever he wanted, and he made it clear he wanted to return home to Los Angeles to play for the Lakers. Presti made a bet on an asset that would probably skip town in a year’s time. In doing so, he risked his job in Oklahoma City, a small market team that doesn’t often lure high-end free agent players.
“Scared money don’t make none,” Presti famously remarked at the press conference, referencing a lyric from Midnight, by A Tribe Called Quest.
Presti’s big bet paid off. George liked it so much in the midwest that he spurned the Lakers and signed on for another year with the Thunder, shocking many.
Montreal’s next tech unicorn?
Like Presti’s gamble, risk-taking is sometimes necessary in life. And for so long it felt like Canadian tech markets like Montreal couldn’t stacked up to Silicon Valley and New York because of a lack of risk.
We’ve heard the same refrain over and over again: Canadian markets lack the big bet behaviour to compete with our American neighbours.
But in the October of 2016, the beginnings of what would become one of Montreal’s biggest bets to date began to take shape. A new incubation-style “AI startup factory” launched in an attempt to retain the city’s AI talent and provide a spark for young companies.
The company was named Element AI.
Four years later as we examine the wreckage, the question remains that has polarized Montreal’s tech community. Was Element AI positive or negative for Montreal’s tech sector?
Hype and fanfare
Element AI burst onto the Canadian tech scene four years ago amid hype and fanfare, promising to deliver AI-powered operational improvements to a range of industries and anchor a thriving domestic AI sector, wrote the Globe and Mail at the time.
“The company became the self-appointed representative of Canada’s AI sector, lobbying politicians and officials and landing numerous photo ops with them, including Prime Minister Justin Trudeau.”
But the mega-funded Montreal-based AI research company sold to the Silicon Valley cloud computing giant ServiceNow just before Christmas, 2020. The bargain-bin price tag was US $230 million, and adjustments and expenses could bring the final price down to US$195 million. This was far below Element AI’s reported value at between $750 million and $880 million when it last raised money ($200 million at the time) in September 2019.
Debate abounded following the sale. How could investors pump so much public and private money into Element AI only to see this ending? How could it get to a point where actual earnings dramatically trailed overhead?
I spoke to several Montreal-based AI startup founders about this story. Some see it as a disappointing tale of a community desperately trying to build its own flagship version of tech success.
They’re frustrated that smaller AI startups in Montreal were ignored by local public and private capital firms. They say investors were too enamoured with Element AI, leaving little capital or facetime for smaller companies desperately competing for a slice of funding.
Others say the city is better off after the tale of Element AI, and that investors are now being chastised for taking the swing-for-the-fences gamble that many had long pushed for.
They say it finally brought legitimacy to a city that had long possessed both the talent and AI research: Montreal sealed its reputation as the AI hotbed.
‘Win at all cost’ funding mentality
One aspect to Element AI’s story was the amount of public-private funding it took on, most of which was, indirectly, taxpayer money. Governmental funding bodies remain among the largest contributors to the majority of venture funds in Quebec.
Over four years, Element AI took on roughly $342 million in investment. Most of that was venture capital while $5 million was debt financing from the Government of Canada.
Dr. Jason Behrmann questions why the “grow fast at all costs using VC funds” model remains the most popular. Behrmann is the Director of Marketing & Communications at Zetane, a Montreal-based AI startup that provides intuitive software for exploring, debugging, and perfecting machine learning models.
While focusing heavily on fundraising, Element AI took two years to focus on product development after initially pursuing consulting gigs. The firm “struggled to advance proofs-of-concept work to marketable products.” Several client partnerships faltered in 2019 and 2020, reported the Globe and Mail.
Meanwhile, hype and international renown grew, in part due to Element AI’s association with star AI and machine learning professor, Dr. Yoshua Bengio.
Behrmann doesn’t understand how the hyped expectations grew out of control so fast.
He says government funding organizations have built-in vetting processes before funding companies. Those checks and balances are supposed to ensure public tax dollars go towards promising companies proposing realistic projects.
AI community saw it coming
Every company has their strengths and weaknesses, said Behrmann. One of Element AI’s strengths was strong brand recognition, great scientific researchers and leaders with proven track records working in tech and investing. Everything on paper screamed success.
Yet what could have been over-confidence and blind trust among public and private investors somehow led to an 18-wheeler going right off a cliff.
Behrmann knew several people at Element AI and called the story “disheartening”. He said the AI community saw it coming for over a year now. There were rumours of dysfunction within the company.
“It’s not so much whether or not you take funding from here or there. It’s more do you have the right foundation before asking for money,” said Behrmann. “I don’t know why they overlooked the fact that they did not have a solid product, no traction, and no real client base.”
Meanwhile, he said, companies like Zetane and others with real products and customers struggled to gain the attention of Montreal-based investors.
Tough time gaining the attention of investors
If Behrmann is the pleasantly-spoken marketing professional on one end of the spectrum, Guillaume Bouchard is the straight-talking CEO on the other end.
Bouchard has been running businesses for 21 years. He once was the youngest North American franchisee manager while he worked for College Pro Painters during summers while studying at HEC. Bouchard said he went on to lead the largest digital marketing agency in Canada from 2005 to 2017, NVI, which was later sold to Dentsu in 2013.
After that he spent time at Helge Seetzen’s TandemLaunch, the notable Montreal-based company-creator.
Despite the nice resume, Bouchard couldn’t get a sniff from Montreal investors in 2018 while pitching his AI company, CONTXTFUL. He said his team met with 70 to 75 different funding sources before finally catching the attention of a Quebec City-based firm, Innovexport.
Bouchard had always been taught that in order to seek investment money, one needs to show what they’re worth. Show them that you’ve actually built something.
Along with CONTXTFUL, Bouchard said several AI startups faced difficulties accessing critical funding during this period.
“When you take $200, $300 million and you put it into one business, the reality is that there are probably 50 smaller AI shops that didn’t get $3, $4, $5 million in seed funding,” said Bouchard. He’s referencing the $137 million funding round which Element AI received in 2017 and the $200 million it received in 2019.
Once considered the number one destination in the world to work in AI, Montreal’s tech community “lost some of its mojo” after the Element sale, said the tech CEO.
“Especially the employees that were burned, lost their jobs and left their home to live here in the hopes of building something great,” said Bouchard. “That’s the big thing for me, that the acqui-hire approach was more important than what it was trying to accomplish as an AI product company.”
Moreover, Behrmann and Bouchard say Element AI paid inflated salaries to some staff which hampered the ability of smaller companies to compete for that same talent.
Bouchard said smaller AI shops were penalized. They simply couldn’t compete with Element AI to hire data scientists earning six-figure salaries. It became what he called an “IT vacuum,” where all the best candidates were mainly going to one large AI startup.
“There were a lot of people who left their careers internationally, landed in Montreal, and now they’re used to earning $2 or $300,000 a year,” said Bouchard. “How motivated are they going to be when they get a job for half that amount? Are they going to stick around in two or three years or just go back home?”
Behrmann hopes the aftermath will bring balance and realistic price adjustments to talent.
Credits and grants won’t be recouped
Some suggest the notion that the government recouped its investment in Element AI from the US $230 million sale price doesn’t paint an accurate picture.
A source preferring to remain unnamed believes whatever amount Element AI took in SRED credits won’t be recouped by Quebec’s government.
“It has caused the taxpayers so much money in credits and grants,” said the source. “Often we forget that there’s not just the investment itself, but also all the subsidies, grants and tax credits.”
That source said they want to see some sort of a vision or plan from Quebec’s government and the VC funds involved in the Element deal.
“I haven’t seen anyone from the government explaining how they’ll reuse that money and we haven’t seen any VCs do any sort of a mea culpa. These guys are just not talking about the deal,” said the source. “I think we’re long overdue for an answer.”
Bouchard, meanwhile, wonders how the government will create roles for small, medium and large AI firms. That is, he says, so we don’t make the same mistake twice.
“How can we cut $5 or $10 million dollar cheques and create some sort of a better, stronger pool of vertically-focused AI shops creating long term value for the economy here in Quebec?”
Invest-AI gives financial support to companies willing to integrate AI into their processes, by reimbursing partial project costs. This means that smaller AI shops working with those companies gain business. It’s co-investment program will match project costs, dollar for dollar, up to $1 million.
Meanwhile, Scale AI is a government-powered fund that invests in Canadian companies using AI to power their supply chain, and also invests in smaller AI shops to help finance their growth.
Element AI rose the tide in Montreal
For Mark Maclean, an executive at Montreal International, Montreal’s tech community gained significant value from Element AI. Maclean’s job at Montreal International is to attract foreign companies to set up a presence in Montreal and Quebec.
He feels some are being overly negative about Element AI.
“Rising tides raise all boats and I think Element AI rose the tide in Montreal,” said Maclean. “It gave Montreal notoriety for other companies to benefited from in the short term and over the long term.”
Maclean brought several companies to meet Element AI during recruitment visits to Montreal. Similarly, his job entailed recruiting some of Element AI’s competitors to set up shop in Quebec as well.
It was a positive to be able to tell prospective companies that Montreal was raising the most venture capital in the country. This was in large part due to Element AI.
“Element AI tried to raise significant money and do it the Silicon Valley way, which happens 10 times a month in the Bay area. We tried it here and it didn’t work out. But we need to take risks, right?” said Maclean.
“It was a big bet and the government investors more or less broke even. The ecosystem built and learned a lot from this and there will be startups spinning out of this.”
What transpired is not the disaster that it’s being made out to be, he added.
“It’s not the worst outcome. It was good for Montreal and I think it will continue to be good. Servicenow is an extremely well-respected company, and this validated how good the Montreal ecosystem is,” said Maclean.
Where do we go from here?
Moving forward, Behrmann hopes investors will take the example of Element AI as a wakeup call.
“I hope the government and other major funding bodies learn that a better way forward is to invest in companies with at least a provable product on the market for a little while,” he said. “It would be irresponsible to invest in companies otherwise.”
The unnamed source took that sentiment a step further.
“How the money flows in Quebec is broken, and Element AI did not help to fix it,” they said. “That’s the sad part, because sometimes it flows not through potential, but through politics.”