A company can become excited about artificial intelligence (AI) while still refusing to sign a cheque. The technology needs to solve something expensive, slow, or frustrating before it deserves a permanent place in the budget.
That’s a healthy reality check for investors too. Spending on computing equipment doesn’t tell us whether customers will earn enough from the results. I prefer a business that helps clients put the technology to work and gets paid for delivering something useful.
Renewed concern about AI spending continues to weigh on technology stocks. That said, it makes a quieter part of the sector more interesting. Companies helping customers lower costs can have a reason to win work even when enthusiasm for ambitious experiments cools.
Source: Getty Images
Start with the bill
A bank doesn’t need another impressive demonstration if its employees still spend hours moving information between systems. A government department doesn’t gain much from a clever chatbot that can’t safely access the records required to answer a request.
Useful projects connect technology to work that needs doing. That can involve updating old software, organizing data, improving security, or taking over ongoing operations. AI may be one component of the improvement, instead of the whole product being sold.
The customer’s math should be straightforward. A project costing $2 million needs enough savings or additional earnings to justify that spending. If it saves $1 million annually after ongoing costs, it recovers the initial cost in about two years. That illustration explains the commercial hurdle, not the terms of a particular contract.
CGI
CGI (TSX: GIB.A) provides technology and business consulting services, including systems integration and managed operations. Its work connects it to clients who need existing systems to function while adopting newer tools.
Management’s July results discussion emphasized “modern data, cybersecurity and organizational readiness.” Those are practical requirements. A client can’t confidently automate important work if the underlying information is unreliable or the controls aren’t ready.
CGI’s fiscal third-quarter revenue was $4.2 billion. Operating cash flow reached $604.5 million, roughly $0.14 for each revenue dollar. That cash helps support the business while clients decide which new projects to approve.
The company also reported $31.8 billion of contracted backlog. Backlog is work expected under agreements, not cash already collected. It nevertheless gives CGI a substantial base beyond the next wave of AI announcements.
Considerations
Adjusted earnings per share (EPS) rose 9% to $2.29, while revenue increased 2.5%. Profit per share can benefit from operating decisions and a changing share count as well as sales growth. I like the combination of established revenue and cash generation more than a story relying entirely on future adoption.
I would buy gradually for long-term growth. CGI’s attraction is delivering useful services profitably, not becoming the most talked-about AI name. The dividend is secondary to the earnings and reinvestment opportunity.
Competition remains significant. Clients can delay projects, demand savings, or use new tools to perform some work themselves. CGI needs to adapt its pricing and delivery so that helping customers become more efficient also remains worthwhile for shareholders.
That makes it a different holding from a high-yield dividend stock. It can bring another earnings source to a Canadian investment portfolio, with exposure to customers and services beyond resource prices and domestic lending.
Bottom line
I’d choose CGI as a measured way to invest in businesses making technology useful. A more demanding customer can actually strengthen the appeal of experienced implementation help. The opportunity is to earn from work clients can justify paying for, with cash generation supporting the case while the market argues about how much AI spending is enough.
