Greg Wilkes, founder of Develop Coaching, lays out pragmatic ways small and medium-sized construction firms in the UK can use Artificial Intelligence to protect margins and lift productivity. He compares the shift to swapping a hand saw for a power saw: the job is the same, but the rhythm, speed and finish change. This is not about sweeping enterprise overhauls; it is about targeted tools that plug into what teams already do and remove repetitive drudge.
The recommended approach is intentionally simple. Map one painful workflow on a whiteboard, identify the repetitive or data-heavy steps, then pilot a small tool that addresses that bottleneck. Practical pilots can be run over a weekend with usage-based pricing. The article highlights five use cases that typically pay for themselves quickly: estimating assistance trained on historic quotes and supplier rates; predictive scheduling that re-sequences trades before problems cascade; contextual document search that finds the right clause in seconds; computer-vision safety monitoring to nudge supervisors and build evidence for clients and insurers; and invoice capture paired with forecasting to see cash flow ninety days ahead.
Each use case is described with a focus on measurable outcomes. Estimating reduces nasty surprises. Scheduling prevents last-minute firefights. Document assistants cut interruptions. Vision systems lower incidents and create shareable records. Invoice automation frees time and uncovers impending cash squeezes. All are presented as add-ons rather than locked-in systems, with pilots using existing photos, schedules and accounts data.
Wilkes outlines a three-month test-and-scale rhythm. month one: choose a workflow and define success in plain language. month two: connect the pilot to existing systems and appoint two internal champions, one site-focused and one numerate. month three: measure, decide and either scale or stop. The aim is a steady cadence of small wins, reviewed weekly, that compound into meaningful operational change.
There are governance pointers too. Set two non-negotiable profit rails around gross and net profit, keep a dated ´model of record´ for pricing and safety outputs, and store sensitive client data in secure repositories rather than random tools. Accountability should focus on a tiny set of measures that lead to action, not dashboard obsession.
The final message is clear: smaller contractors will out-learn, not outspend, larger rivals. Document one process, pilot one tool, review weekly and repeat. In construction, incremental advantages stack up; that steady compounding is what builds lasting operational resilience and a calmer cash position.
