Licence-and-pray is over: Boxtree bets Lean methodology can fix UK AI's ROI problem

The problem hiding behind every Copilot rollout

The script is familiar by now. A CEO announces the firm is "embracing AI". Fifteen Copilot licences get purchased. A few people write better emails. Marketing copy sharpens up. And then, six months later, someone asks where the returns are.

Boxtree Consulting, founded by Max Pardo-Roques from Birmingham's Jewellery Quarter, was built specifically to answer that question. Launched in July 2026, the firm applies Lean methodology, derived from the Toyota Production System, to locate the manual processes AI can automate, from invoice checking and data entry to scheduling and compliance, and then builds solutions inside the systems clients already run. No new ERP. No bolt-on platform. Just measurable waste reduction.

"Most of the industry has this backwards. It starts with the technology and goes hunting for a use case," Pardo-Roques told Pressat. "We start with the process and the value attached to it, and we only automate where the numbers stack up."

The firm guarantees every engagement is cost-neutral within its first year, targeting a five-to-one return on investment.

Why now, and why Birmingham

Pardo-Roques spent more than 15 years in operational excellence, most recently as commercial director at Develop Consulting, working with clients across manufacturing, FMCG, airports and logistics. The pivot to AI is less a reinvention than a natural extension: Lean is, at its core, a discipline for finding and eliminating waste. AI, deployed against the right process, is simply a faster eliminator.

The timing reflects a structural problem in the UK market. Fewer than three in ten UK businesses with under 50 staff were using AI in 2025, against 44% of firms employing more than 250, according to ONS data analysed by the Bennett School of Public Policy. That gap between enterprise adoption and the SME base is where Boxtree positions itself.

Boxtree's diagnostic model is deliberately unglamorous: map processes, put a pound value on the waste within them, build a business case for each opportunity, and only proceed where first-year returns provably exceed the cost of the work. It is the kind of rigour that most vendor-led implementations skip entirely.

The talent crunch underneath the spend

Boxtree's launch is also well-timed relative to a wider supply crisis. Research from freelancer platform Malt, published earlier this month on Consultancy.uk, found that demand for AI agent expertise increased by 5,800% over the past year across Europe, while the AI Engineer community on Malt's platform of one million freelancers grew by 229% in the same period. AI is now the second most in-demand skill on Malt's platform regardless of company size or sector.

The implications for consulting practices are direct. Firms that can deliver AI outcomes without requiring clients to hire scarce specialist talent in-house, and without billing eye-watering day rates for that talent, hold a structural advantage. Boxtree's model, embedding solutions into existing systems rather than deploying fresh infrastructure, sidesteps the worst of the procurement and integration overhead.

ManpowerGroup's 2026 Talent Shortage Survey adds a further data point: 73% of UK employers report significant hiring difficulty, and AI skills have claimed the top spot globally for hiring difficulty for the first time in the survey's history.

What the company formation data says

Boxtree is not alone in spotting the opportunity. AI Business Dispatch analysis of Companies House data (as of July 2026) recorded 2,623 new SIC 70.22 (management consultancy) incorporations in Q3 2026. That figure represents a sharp contraction of 79.5% against the prior period, suggesting the initial wave of AI-adjacent consultancy formation has passed and the cohort now entering is smaller and, arguably, more purposeful.

The brand picture is starker still. UK IPO trademark data shows just 473 Class 45 filings in Q3 2026, down 70.2% versus the prior period. More revealingly, 99.7% of active SIC 70.22 companies hold no Class 45 trademark at all, per AIBD analysis of Companies House and IPO data as of July 2026. For a sector selling credibility, that level of brand under-protection is remarkable. Most new consulting entrants are, in effect, operating with no formal IP shield on their service identity.

The uncomfortable arithmetic of utilisation

The broader context for Boxtree's pitch is the utilisation problem that Thomson Reuters' 2026 AI in Professional Services Report captures crisply: organisation-wide AI usage has almost doubled to 40% in the past year, yet a clear gap is emerging between organisations achieving real impact and those falling behind.

Dayshape's analysis of the same dynamic is worth sitting with. AI saves time, but not always in neat, reusable blocks. In service lines where work arrives in fixed chunks, shaving hours off a task does not automatically free capacity for another engagement. The firms that benefit most will be those that can deliberately turn fragmented time savings into usable capacity across teams.

Boxtree's Lean-first framing applies the same argument to clients rather than to the consultancy itself: productivity gains at the individual level do not become business value unless someone has deliberately redesigned the process around them.

The question for any consulting firm watching this launch is not whether the approach is sound. It is whether the mid-market client base is now ready to pay for process discipline rather than technology novelty. Given that 73% of UK employers are still struggling to hire, and that nearly 85% of AI projects fail to escape the pilot stage according to industry benchmarks, the answer is probably yes. Whether Boxtree or someone else gets there first is another matter.