q.beyond is turning the efficiency promise of artificial intelligence into a restructuring programme, with the German IT services company preparing job cuts, moving more support work to Romania, and taking up to €6 million of one-off costs after internal AI use exposed larger savings than management had expected.

The Cologne-headquartered company said wider deployment of AI across its operations had revealed enough additional efficiency to accelerate existing transformation plans. Managed services and administrative roles will be affected, while q.beyond’s recently established site in Cluj will take on AI-assisted round-the-clock customer support from the third quarter.

Unlike corporate AI programmes framed mainly around software access or productivity aspirations, q.beyond has attached the technology directly to its cost base. The accelerated programme is expected to cost between €5 million and €6 million during 2026, but management forecasts annual savings of roughly €7 million from 2027, including more than €1 million from the move towards AI-supported service delivery in Cluj.

The company has already booked a €0.9 million provision in the second quarter for reorganising its service operation. Positions will be removed over the coming months, particularly in managed services and administration, although q.beyond has not disclosed the total number of roles affected.

Those changes arrive alongside softer trading conditions among the company’s core German mid-market customers. Second-quarter revenue fell to €43 million from €44.4 million a year earlier, while adjusted EBITDA before the initial restructuring provision was €2.5 million, compared with €2.7 million during the same period of 2025.

As transformation costs combine with continued reluctance among German mid-sized businesses to invest, q.beyond has lowered its full-year revenue forecast to between €176 million and €180 million from a previous range of €182 million to €190 million. Its EBITDA outlook has fallen more sharply, from €10 million–€16 million to €3 million–€7 million, and the company now expects both a negative consolidated result and negative free cash flow for 2026.

The restructuring provides a more concrete test of enterprise AI than another announcement about employees receiving access to assistants. q.beyond is effectively betting that automation can remove enough operating cost to repay a sizeable restructuring charge within roughly a year, while maintaining the service levels expected from a provider responsible for cloud systems, SAP and Microsoft applications, security, and other business-critical technology.

AI adoption is also becoming entwined with the geography of service delivery. Moving 24/7 support activity to Romania would already change the economics of the operation; adding AI-assisted processes potentially changes the number and type of people required to run it. When savings begin appearing in q.beyond’s accounts, the effects of automation and lower-cost delivery will be difficult to separate.

A similar tension between automation and employment is appearing elsewhere in European technology and communications businesses. Telefónica Deutschland is cutting up to 1,100 roles while pursuing AI-related efficiency savings, as companies move from treating generative AI as an additional workplace tool towards redesigning established functions around it.

For q.beyond, the technology sits on both sides of the commercial model. The company is applying AI to lower its own operating costs while simultaneously trying to sell AI-enabled services to customers, including automated business processes and managed AI workflows. Internal deployment therefore becomes part of the sales case: if the technology materially improves its own managed-services economics, q.beyond has stronger implementation evidence to take into the German mid-market.

Execution remains difficult, particularly in customer support where the work cannot be reduced to a queue of tickets waiting to be automated. Savings depend on whether automated processes resolve problems reliably, whether unusual incidents reach skilled staff quickly, and whether service quality survives the reorganisation rather than merely becoming cheaper to provide.

The weaker investment environment adds another complication because improvements in operating efficiency could initially be obscured by softer revenue. q.beyond is accelerating its AI restructuring while customers remain cautious about technology spending, leaving management to deliver cost reductions without assuming that growth will cover mistakes in the transition.

The first useful evidence should emerge during 2027, when the restructuring costs have passed through the accounts and the company expects the €7 million of annual savings to begin supporting profit and cash generation. By putting unusually specific numbers around AI-led operating change, q.beyond has made its subsequent financial performance a test of whether the efficiency promised by automation survives inside a real service organisation.