Teleperformance

Artificial intelligence is challenging traditional outsourcing revenues while creating demand for new enterprise services. Teleperformance's latest results show both sides of that transition.

For the world's largest customer experience providers, AI presents a dilemma: technology capable of automating customer interactions can also create opportunities to deliver more sophisticated services. Teleperformance, operating under the TP brand, is navigating precisely this tension.

On 5 November, the French group revised its 2026 revenue outlook to a like-for-like decline of 1.5% to 2.5%, against its previous expectation of flat to 2% growth. Shares fell 15.3% the following day. Yet its third-quarter figures reveal differing trajectories across the business.

Beyond the Headline Decline

Group revenue fell 1.2% like-for-like during the quarter. Excluding Trust & Safety, where content-moderation volumes are under pressure from automation, revenue grew 2%. Core Services excluding that activity expanded by 2.9%.

These figures do not negate the decline in Trust & Safety, an important part of TP's portfolio. They do, however, show that the pressure is not uniform.

The wider outsourcing market faces a similar structural question: how quickly can growing demand for complex, technology-enabled services offset the automation of routine work?

TP has disclosed approximately 1,400 AI projects at the end of September 2026, more than double the previous year's level. Over 300 new deployments entered production during the third quarter. More than 35% of its clients now use at least one TP.ai solution, while the company says over 70% of new deals won during the year include a TP.ai component.

The figures suggest AI is becoming embedded in TP's commercial offering, rather than remaining confined to experimental pilots. But adoption is not the same as revenue: the company has yet to demonstrate how much these deployments contribute to sustainable growth and earnings.

Humans Remain Part of the Model

TP is positioning its services around a combination of automation and human expertise. Automated systems can handle repetitive tasks; specialist teams remain necessary for complex interactions, oversight, quality assurance and sensitive decisions.

The ambition extends beyond customer-facing tools. Bloomberg reported in July on chief executive Jorge Amar's plan to make TP's workforce AI-enabled by early 2027, signalling an effort to integrate the technology into everyday service delivery. For enterprise customers, the commercial test will be whether this combination improves efficiency and service quality without sacrificing accountability.

Despite cutting its revenue outlook, TP maintained its 2026 recurring EBITA margin target of around 14.6% and net free cash flow guidance of €800 million to €850 million, excluding non-recurring outflows. It also expects restructuring measures to deliver €150 million to €170 million in annualised savings.

Those targets provide financial room for transition, although restructuring costs and uncertainty around legacy volumes remain significant. TP plans to update its 2028 financial objectives in the first half of 2027.

The central question is therefore not whether AI will change outsourcing: it already is. It is whether TP can turn growing adoption of its AI-enabled services into enough profitable business to offset contraction elsewhere.

Its latest figures suggest meaningful commercial progress, but the next stages of execution will determine the outcome.