Startups enrolled in accelerators and incubators across Canada generally earned less revenue than firms that didn’t participate, according to a recent report from the department of Innovation, Science and Economic Development. 

The data, based on a Statistics Canada analysis, stands out in an overall rosy report boasting of the benefits of Canada’s accelerator programs. “There is early evidence that [business accelerators and incubators] are associated with the growth trajectory of high-potential Canadian startups,” the report reads, adding that “governments are keen to sustain” such companies, “and see them scale to become future Canadian anchor companies.”

The report is meant to shed light on the effectiveness of startup-assistance programs, many of which are heavily subsidized by the federal and provincial governments. It comes as high-profile accelerators in Canada grapple with their place in the innovation ecosystem, including Techstars Toronto ending its programming, and restructuring at MaRS Discovery District. 

The report looked at about 8,000 survey responses collected by business accelerators and incubators (BAIs) from startups that participated in their programs from 2017 to 2020. The information was cross-referenced with tax-filing data from Statistics Canada. The innovation department then analyzed the findings and compared them with a group of companies that had no links to BAIs. 

Startups in the 2020 cohort reported higher revenue each year from 2016 to 2020, while revenue for the control-group remained fairly stable. Still, BAI startups generally underperformed other companies, with nearly 30 per cent less revenue on average over the study period. The highest revenue-generating startups were the exception, posting slightly more than non-BAI firms, but only in 2020. 

The report attributed the lower revenue in part to BAI startups being typically very young, with nearly 70 per cent of firms less than five years old, in contrast to about 73 per cent of comparison companies being older than 16 years: “Therefore, it is not surprising that the…companies exhibited lower revenue figures than the comparison group from 2016 to 2020.”

After controlling for factors like the company’s age, industry, region and number of employees, accelerator- and incubator-supported startups had revenue that was 13 per cent higher than other firms. But that advantage was short-lived, with startups generating no more revenue than a typical company within a year after participating in an accelerator or incubator program. The revenue reflects startups’ earnings from core business activities, ISED spokesperson Justin Simard told The Logic by email, and doesn’t include investments from accelerators and incubators.

The value that accelerators and incubators bring to participating companies and to the broader economy is still not clear, said former BlackBerry co-CEO Jim Balsillie. “Beside the obvious fact that ISED should not be evaluating the effectiveness of their own programs, this report conflates correlation with causality,” said Balsillie, a frequent critic of accelerators. 

The report’s authors, Manasi Joshi and Jiong Tu, were not available for an interview, ISED told The Logic. 

Startups in accelerators and incubators were far more likely to carry out research and development. In 2020, 21.7 per cent of BAI companies spent money on R&D compared to 1.3 per cent of other companies. That number was even higher in Ontario (30.6 per cent) and Quebec (26.4 per cent). 

BAI companies in the 2020 cohort also had higher employment growth compared to other firms, despite the overall number of employees between groups being roughly the same. The BAI startups paid higher salaries, with the average pay increasing for every cohort from 2016 to 2020, while the comparison group’s salaries remained fairly stable. 

Angela Larraguibel, vice-president of founder services at Communitech—a startup hub in Kitchener-Waterloo, Ont., that participated in the study—said the findings reflect the value the organization brings to the startup ecosystem with supports and services that are “critical to helping founders start, grow and succeed.” 

But Balsillie said the report makes no mention of accelerators’ “troubled relationships in the broader innovation ecosystem,” and serves to rationalize Ottawa’s funding for the organizations. “This is a justification for ISED to keep funding for the incubators and [accelerators],” he said.

Some of Canada’s largest accelerators are facing mounting criticism, questioning the value they create from taxpayer dollars that support them. Communitech’s CEO Chris Albinson has come under fire for expanding the organization’s mandate beyond supporting only Kitchener-Waterloo startups. MaRS, a pre-eminent accelerator in downtown Toronto, cut about 20 positions in June, including several senior roles, as its new CEO Alison Nankivell seeks to reset the business model “toward a more agile and lean organizational structure.” The layoffs came less than a month after Benjamin Bergen, president of the Council for Canadian Innovators, derided MaRS and its ilk on stage at the National Angel Capital Organization summit in Ottawa, arguing that the accelerators are failing to turn the public dollars they receive into wealth for Canada.

MaRS did not submit data for the 2020 cohort of the report and declined The Logic’s request for comment.  

ISED’s study acknowledged some limitations, including the short period over which it tracked startups. “Future research will examine additional years following BAI graduation to understand whether there are long-term effects,” the report reads. There’s also potential selection bias, in which highly motivated startups choose to enrol in accelerators and incubators. Those firms may have had strong employment and revenue, for example, even without support from the programs. “Due to a lack of variables on a company’s motivation,” reads the report, “it is difficult to apply a correction method to mitigate the bias.” 

The federal government first announced plans to study the impacts of accelerators and incubators in the 2016 federal budget, and carried out a multi-year pilot phase before publishing the recent report. ISED did not say when it will publish the next report. Researchers are analyzing data for 2021, and Statistics Canada has collected information for 2022. There will be a call for 2023 submissions in the coming months, said spokesperson Andréa Daigle. 

When asked about the data being four years old—from a time when the startup market was much different than it is today—Simard said the results should still hold up. “The findings reveal a positive correlation between BAI support and business performance. It is reasonable to expect that this correlation would hold in recent years,” said Simard. “The continued collection of these data will allow further analysis in the future on this and other questions to better understand the effect of BAI support on their client businesses.”