IT outsourcing was born as an arbitrage: the same work, done competently, for less money somewhere else. That founding logic is fading. The contracts analysed on our technology strategy pages show buyers choosing partners for scarcer currencies — artificial intelligence expertise, cybersecurity depth, and the speed to ship products before competitors — with price demoted to a qualifying round rather than the final.
The change is visible in the deals. Engagements are shorter, scoped around outcomes rather than headcounts, and increasingly co-located in hybrid patterns: discovery and design near the client, engineering in delivery centres, everything stitched together by collaboration tooling that the pandemic years made unremarkable. Clients report keeping the crown jewels — data architecture, security strategy, product direction — in-house, while partners industrialise everything around them.
The AI clause in every contract
Artificial intelligence has rewritten the negotiations twice over. First as a capability clients demand: partners must bring working AI delivery experience, not slideware. Then as a productivity question: if the vendor’s tools make its engineers faster, who keeps the saving? New contracts increasingly answer in writing, with shared-productivity clauses that would have been unthinkable in the day-rate era.
Industry analysts at firms such as Gartner describe the mature market as capability sourcing — buying access to skills that cannot be hired fast enough locally, with cost as table stakes rather than the prize.
What buyers get wrong now
The failure modes have updated too. Choosing a partner for a demo rather than a delivery record. Outsourcing a mess and expecting it to return as order. Underinvesting in the client’s own product and vendor-management talent, then blaming the partner for drift. The successful programmes share a discipline: small number of strategic partners, transparent metrics, and collaboration treated as an operating model rather than a procurement category.
Geography is being redrawn again
The map of delivery is shifting a second time. Clients now blend nearshore teams in their own time zones for collaborative work with offshore centres for scale — a portfolio of locations chosen per function rather than a single bet. Political risk, once a footnote, is a scored criterion: concentration in any one country is treated as a vulnerability to be hedged like currency. Meanwhile the destination countries climb the value ladder — India’s global capability centres now house product ownership and research that headquarters once kept, and Eastern Europe and Latin America compete on engineering depth and language. The industry’s founding question — where is it cheapest? — has been fully replaced by a portfolio question no spreadsheet settles alone: where is each kind of work done best, safest and fastest, all at once?
Outsourcing’s first era asked where work could be done cheapest. Its second asks where it can be done best, fastest and safest — and the answer, increasingly, is a partnership that looks less like a vendor contract and more like an extension of the company itself.
