OpenText and the Waterloo Model: Canada’s Original Enterprise Software Exporter

OpenText is the original proof that a Canadian company can sell software to the world’s largest enterprises, and its last three years are a cautionary tale about what happens when a serial acquirer buys one company too many. From its 1991 founding in Waterloo to the Micro Focus deal that nearly doubled its size, the company has been shaped by acquisitions. This article explains the Waterloo model that produced OpenText, BlackBerry, Descartes and Kinaxis, examines the debt and divestitures that followed Micro Focus, assesses the leadership change and Antoun’s first results, and asks what the region’s enterprise-software cluster looks like after the BlackBerry era. It is part of the Canada Company Stories hub.

How did a dictionary project become Canada’s enterprise software exporter?

OpenText began as a search engine for the Oxford English Dictionary. In the mid-1980s University of Waterloo professors Frank Tompa and Gaston Gonnet developed text-indexing technology for the dictionary’s first computerised edition, and in 1991 they founded Open Text Corporation with Anouar Namouh to commercialise it. The company briefly ran one of the first web search engines before deciding that enterprise document management, not consumer search, was the business to be in.

Tom Jenkins joined as chief operating officer in 1994 and ran the company as chief executive through its formative years, listing it on Nasdaq and the TSX and establishing the acquisition-led strategy that defines it. OpenText bought Hummingbird for US$489m in 2006, GXS, the business-to-business network, in 2013, Dell EMC’s Documentum content-management division for about US$1.6bn in early 2017 and Carbonite, a backup vendor, for about US$1.45bn in 2019. Each deal added a mature product line with a loyal installed base and recurring maintenance revenue.

The logic was financial as much as technological. Enterprise content management, the discipline of storing, governing and retrieving documents, is sticky: once a bank or a government has loaded decades of records into a system, migration is expensive and risky. OpenText bought those installed bases at modest multiples, cut costs, cross-sold its own products and used the cash flow to borrow for the next deal. By 2022 revenue had reached about US$3.5bn and the company employed well over 20,000 people.

Mark Barrenechea, an American executive who had run Silicon Graphics, became chief executive in 2012 and added the chief technology officer title in 2016. He accelerated the pace of acquisitions and adopted the “Information Management” framing that the company still uses. Jenkins remained as chairman and, as events in 2025 would show, as the ultimate authority in the boardroom.

Why did OpenText buy Micro Focus, and what did it cost?

OpenText bought Micro Focus because it was the largest available pool of mature enterprise-software revenue, and it cost far more than the headline price. Announced in August 2022 and completed in January 2023, the acquisition of the UK-listed group was valued at about US$5.8bn to US$6bn including debt. Micro Focus itself was the product of a disastrous 2017 merger with Hewlett Packard Enterprise’s software division and had been shrinking for years; OpenText argued it could apply its integration playbook to stabilise the business.

The deal roughly doubled OpenText’s revenue, to about US$5.8bn in fiscal 2024, but it also roughly doubled its debt and brought in dozens of overlapping products with separate codebases and sales teams. Analysts such as Omdia’s Rik Turner later noted that running two large, un-integrated portfolios was expensive and complex, and the promised cross-selling was slow to materialise. Rising interest rates in 2023 made the acquisition debt more costly to carry.

The first retreat came within eighteen months. In May 2024 OpenText sold the Application Modernization and Connectivity business, the mainframe software that had been Micro Focus’s crown jewel, to Rocket Software for about US$2.28bn, using the proceeds to pay down debt. The sale removed a high-margin, slowly declining business that private-equity-backed Rocket valued more highly than public-market investors valued it inside OpenText.

The remaining Micro Focus assets, including the Vertica analytics database and various security and testing tools, continued to shrink. Fiscal 2025, ending 30 June 2025, brought total revenue of US$5.17bn, down 10.4% as the AMC divestiture and organic declines compounded. Non-GAAP earnings per share fell 8.4% to US$3.82. Headcount reductions ran alongside: about 2,800 positions were eliminated from 2024, including 1,600 cuts announced in May 2025, as part of a three-year programme targeting roughly US$500m of run-rate savings.

Why did the board remove Mark Barrenechea?

The board removed Barrenechea because the Micro Focus strategy had not produced the growth or the share-price performance that justified it. On Monday 11 August 2025, four days after the disappointing fiscal 2025 results, OpenText announced that Barrenechea had departed as chief executive, chief technology officer and vice-chairman, effective immediately. James McGourlay, a 28-year veteran who ran international sales, became interim chief executive, and Savinay Berry was promoted to chief technology officer, ending the combined CEO-CTO role.

Executive chairman Tom Jenkins thanked Barrenechea for 13 years of service and said it was the right time for a transition. No further reason was given, but the context was clear. OpenText’s shares had fallen about 27% over the preceding five years while cloud-native rival Box had risen 85%. National Bank analyst Richard Tse attributed the exit to an inability to surface value from a sprawling product portfolio; others pointed to a highly centralised management style that limited innovation in the divisions.

The board simultaneously announced a strategic review. OpenText would work with financial advisers to refocus on its core “Information Management for AI” business and exit non-core assets, redeploying capital to growth and shareholder returns. On the first-quarter call in November 2025 management said it expected to divest non-core units at a pace of roughly one per quarter over about a year. An analyst cited a target of 15% to 20% of revenue; management did not confirm the figure.

The pattern of a long-serving acquisitive chief executive removed after one deal too many is familiar in Canadian corporate history, and the contrast with Constellation Software, which has never done a transformational acquisition and whose founder left on his own terms, is instructive; see the Constellation Software story.

What has OpenText sold, and what does it still own?

OpenText has sold three units since the review began and kept the content-management core. The eDOCS legal document-management product went to NetDocuments for US$163m, completed on 12 January 2026. The Vertica analytics database, part of the Micro Focus inheritance, was sold to Rocket Software for US$150m, announced in February and completed on 11 May 2026. Net proceeds from divestitures in fiscal 2026 were about US$312m, earmarked for debt reduction.

What remains is organised around content management, business network, cybersecurity, application development and IT operations, with the content-management cloud as the growth engine: it grew 18% in the December 2025 quarter. Annual recurring revenue was US$4.25bn at fiscal year-end, up 1.3%, and enterprise cloud bookings rose 22.5% to US$947m, the most convincing forward indicator in the results.

The divestiture programme is likely to continue. Management has described fiscal 2027 as a “foundation year” and has not given numerical guidance, which leaves room for further portfolio moves. The Micro Focus security and testing businesses, which overlap with products from larger specialist vendors, are the obvious candidates, though the company has not named them.

What do the fiscal 2026 results say about the turnaround?

The fiscal 2026 results say the bleeding has stopped but growth has not returned. Total revenue for the year to 30 June 2026 was US$5.246bn, up 1.5%, the first increase since the AMC sale. Cloud services and subscriptions revenue was US$1.959bn, up 5.5%, with the fourth quarter at US$503m, up 6.0%; the company has now reported more than 20 consecutive quarters of organic cloud growth. Adjusted EBITDA was US$1.903bn, a 36.3% margin, with the fourth quarter at 37.6%.

Cash improved more than revenue. Operating cash flow rose 21.2% to US$1.007bn and free cash flow rose 17.5% to US$808m. GAAP net income attributable to OpenText was US$643m, up 47.5%, and non-GAAP earnings per share rose 15.7% to US$4.42. The balance sheet at year-end showed about US$5.77bn of debt against US$956m of cash, which implies net debt of roughly US$4.8bn, still high for a company with flat revenue but reduced from the post-Micro Focus peak.

Capital returns reached a record US$677m: US$409m of buybacks under a programme raised to US$500m in February 2026, and US$268m of dividends, with the quarterly dividend lifted 5% to US$0.28 per share. A renewed buyback authorisation covers up to 23.8 million shares through August 2027. Chief financial officer Steve Rai named cash generation, debt reduction and capital allocation as the priorities for fiscal 2027.

Who is Ayman Antoun and what is his strategy?

Ayman Antoun is a 35-year IBM veteran who was hired to turn OpenText from a collector of software assets into a focused data company. Announced on 29 January 2026 and in post from 20 April, he had been president of IBM Americas from 2020 to 2023, overseeing the United States, Canada and Latin America, and earlier ran business-market sales at Bell Canada. He sits on the boards of TD Bank and CAE, which gives him an unusually deep network among OpenText’s largest Canadian customers. McGourlay moved to president and chief client officer, and Jenkins returned to the chair.

Antoun’s pitch is that AI has created urgency in every organisation, but that trusted data decides whether AI delivers value, and that OpenText’s decades of enterprise content make it the secure data foundation in the AI stack. The company joined the OECD’s global safe-AI reporting framework in May 2026 as one of the first Canadian firms to do so, expanded sovereign-cloud offerings through the AWS European Sovereign Cloud and a partnership with Google-backed S3NS in France, and in September 2026 announced a partnership with Cohere to combine its data layer with Cohere’s agentic-AI platform, due to reach customers in early 2027. That link to Canada’s AI champion is explored in the Cohere story.

Investment is following the pitch. In June 2026 OpenText announced a €105m programme in Ireland to create 400 jobs in Cork and Galway focused on agentic AI and sovereign cloud. Fiscal 2027, in management’s words, will be about disciplined execution, expanding sales capacity, deepening partner reach and increasing organic investment in the core portfolio, which is a polite way of saying no large acquisitions.

How did OpenText refinance its debt in 2026?

OpenText refinanced its nearest maturity in September and October 2026 to push the debt wall out to the 2030s. On 22 September the company gave conditional notice to redeem all US$1.0bn of its 6.90% senior secured notes due 2027, launched a cash tender for up to US$450m of its 2028 notes, and began marketing a new US$1.0bn senior secured offering. The new notes priced on 23 September in two tranches: US$500m of 6.70% notes due 2031 and US$500m of 7.15% notes due 2033.

The tender was less popular than hoped, and on 25 September the cap was cut to US$300m. The new notes closed on 1 October 2026 and the revolving credit facility’s maturity was extended from December 2028 to October 2031. The exercise did not reduce total debt, but it removed the 2027 refinancing risk and locked in rates that, while higher than the retired 2027 coupon, reflect the company’s improved cash generation.

Net leverage has not been disclosed in the fiscal 2026 release, but the arithmetic of US$4.8bn of net debt against US$1.9bn of adjusted EBITDA implies a ratio around 2.5 times, down from above 3 times after Micro Focus. For a business whose revenue is flat, that is the level at which rating agencies and lenders become comfortable again, and it explains why the board felt able to raise the buyback.

What is the Waterloo model after BlackBerry?

The Waterloo model is a university-anchored cluster that keeps producing enterprise software exporters even after its most famous company collapsed. The University of Waterloo’s co-operative education programme, the largest in the world, sends engineering students into industry for alternating terms, and the university’s intellectual-property policy lets inventors own what they create. OpenText was the first big spin-out; BlackBerry, founded as Research In Motion in 1984, was the second and for a decade the most valuable company in Canada.

BlackBerry’s decline after 2011 scattered thousands of experienced engineers and managers across the region, many of whom founded or joined the next generation of companies. The Communitech hub, the Velocity incubator and the Perimeter Institute gave the cluster infrastructure, and the Toronto-Waterloo corridor is now routinely described as the largest technology cluster in Canada. The ecosystem is mapped in the Toronto entry of the City Startup Atlas.

The companies that survived are mostly in unglamorous enterprise niches. Descartes Systems Group, founded in Waterloo in 1981 and run by Edward Ryan, sells logistics and customs software, has grown through dozens of small acquisitions, employs more than 2,000 people and reported revenue above US$600m in its latest fiscal year. Kinaxis, in Ottawa’s Kanata district, sells supply-chain planning software, reported US$548m of revenue in 2025, rebuffed activist Daventry Group’s 2024 call to sell itself and appointed Razat Gaurav as chief executive in January 2026. Both are capital-light, acquisitive and export-focused, which is the Waterloo template without the Micro Focus mistake.

The region also shows the limits of the model. BlackBerry itself is now a small security and automotive-software company, the corridor’s largest employers are American technology firms’ Canadian offices, and the venture capital that funds early-stage companies still comes mostly from the United States. The manufacturing side of the Ontario economy, from Magna’s auto-parts empire to the electric-vehicle battery plants, now faces the same question of whether Canadian ownership can be sustained.

What can founders and investors learn from OpenText?

The first lesson is that roll-up strategies have a scale limit. OpenText’s small and mid-sized acquisitions worked for two decades because each could be integrated by the existing organisation. Micro Focus was too large, too complex and too far into decline to be absorbed, and the company spent three years and a chief executive unwinding it. Constellation Software, which has never bought anything it could not integrate within its decentralised structure, offers the contrasting model.

The second lesson is about boards. Tom Jenkins’s board acted within four days of the fiscal 2025 results, replaced the chief executive with an insider, launched a strategic review and recruited an outsider with a clear mandate, all within eight months. That is faster than most Canadian boards move and is the reason the company’s cash flow and share price stabilised rather than deteriorated during the transition. Governance questions of a very different kind arise in the Rogers-Shaw story, where a controlling family rather than an independent board decides who runs the company.

The third lesson is that a mature installed base is a platform if it is managed as one. OpenText’s pitch as the trusted data layer for AI is credible precisely because it holds decades of governed enterprise content; the Cohere partnership and the sovereign-cloud expansions are attempts to monetise that position before it decays. Investors should watch enterprise cloud bookings, which grew 22.5% in fiscal 2026, rather than total revenue, which will stay flat while divestitures continue. For founders in Waterloo and beyond, the company remains the proof that a Canadian enterprise-software exporter can reach US$5bn of revenue, and the warning about what it takes to stay there.

Frequently Asked Questions