Rising token costs and the need for speed push UK enterprises away from custom built AI agents
A major geographical split has emerged between how UK and US technology leaders approach their agentic ecosystems, according to new research analysed in the 2026 Jitterbit AI Automation Benchmark Report.
The UK’s Pragmatic Approach: Buying Over Building
When building and deploying AI agent automations, UK IT decision-makers demonstrate a clear preference for efficiency and speed over bespoke development, with 51% of UK organisations opting for pre-built agents to scale operations, compared to 48% who rely on custom-made.
43% of UK IT decision-makers prefer agents that are embedded in pre-built SaaS applications, but only 28% prefer building their own low-code/no-code integration platforms.
The idea that AI agents could fundamentally disrupt the traditional Software-as-a-Service (SaaS) business model, seems to be more of a myth, as only 20% of UK respondents prefer in-house, custom-coded agents vibe-coded. “Off-the shelf” solution was the least preferred option for UK IT decision-makers, with only 8% opting for this as their primary source of AI agents.
The Transatlantic Flip
When compared to US IT decision-makers responses, the trend is completely flipped. In the US, engineering teams are favouring a less traditional development stance with 55% of US respondents preferring custom-built code developed entirely in-house, and only 45% lean toward pre-built agents.
The Token Cost Reality Check
While the US takes a more proactive engineering stance, changing market economics suggest the UK’s preference for pre-built solutions might be the right call in the long term. As the enterprise market grapples with the escalating costs of tokens and heavy large language model (LLM) consumption, the financial reality of custom development is shifting.
Companies are becoming significantly less inclined to maintain bespoke internal builds when rising token fees make custom alternatives prohibitively expensive compared to predictable, vendor-supported models. Because pre-built and off-the-shelf agents optimise token consumption out of the box, they are becoming increasingly attractive financial alternatives.
Despite in-house developer teams achieving record-high levels of coding sophistication, UK businesses are intentionally looking to pre-built options first and foremost.
Speed as a Corporate Strategy
Globally, the data shows a broader, structural movement toward rapid orchestration over slow manual coding with this transition directly feeds into “Speed as a Strategy”. With 81% of organisations planning to increase funding for AI and automation over the next 12 months, enterprises require scalable solutions that deliver immediate value faster than traditional manual development allows.
Bill Conner, President and CEO of Jitterbit, commenting on the findings from the Jitterbit Automation Benchmark Research, stated:
“We are moving past the era of siloed AI experimentation and entering the era of pure orchestration. The market is waking up to the true total cost of ownership of custom AI. Between developer bottlenecks and soaring token costs, building everything from scratch is a luxury few can sustain. The UK’s slight preference for pre-built applications may have previously been considered cautious, but it’s proving to be highly strategic as economic pressures force a shift from development to intelligent integration.”
Related Post: Roger Bailey is now Vice President of Sales EMEA at Jitterbit
Related Post: Jitterbit Appoints Eddie James as Chief Financial Officer.
