How we work · 2026

Monthly retainer. Powered by Forge.

No start and end dates. We embed for as long as makes sense — usually three to twelve months — at a predictable monthly fee. Scope flexes month-by-month. AI does the heavy lifting; we hold the line. The contracts-manager, estimator and document-controller work runs as one service — drafted by AI, approved by the MD, delivered by an operator who's done the work at a real firm. Three salaries' worth of output for roughly 5% of the cost.

A typical month
Review
Build
Ship
Summary
Week 1
Weeks 1–4
As ready
Month-end
The rhythm

Four phases. Repeated every month.

In practice they overlap — review at the start, summary at the end, build and ship through the middle. Alongside the planned month's work, a live stream of urgent support runs in parallel: fires we didn't schedule for don't wait for the next cycle.

Phase 01 · Review

What's worth doing this cycle.

First few days of the month

We start each month with a short call. What changed? What's biting? What's the highest-impact thing your hours should go to? You don't write a brief — we propose, you push back.

Some cycles lean diagnose more than do — month one of a new engagement, or any month where the real progress is in understanding before automating. Less visible output in those months, but the next cycle ships against reality, not assumption.

  • Last month's outcomes vs what we set out to do
  • What's changed in the business, market, or pipeline
  • Highest-impact moves for this cycle, ranked
  • Agreed scope for the month — written down, three lines
Phase 02 · Build

The actual work.

Weeks 1–4 of the month

Forge holds the workspace — your data, your context, your dashboards, the templates and AI tooling we use to handle the repetitive bits. We're not starting from scratch each cycle, which is why your retainer hours go much further than they look on paper.

  • Weekly check-in (15–30 min) so nothing drifts
  • Updates land in your workspace as they happen — read them when you have time, no chasing for status
  • AI handles the repetitive work; we do the judgement
  • Anything urgent slots in mid-cycle. If it bumps planned work, we tell you up-front what it bumps
Phase 03 · Ship

Things go live as they're ready.

Throughout the month, not just at the end

We don't save deliverables for a big month-end reveal. A page, a dashboard, an automation, a tender response — each ships when it's done, with a rollback plan attached.

  • Each deliverable lives somewhere documented (Forge, your tools, or both)
  • Rollback plan where rollback is possible. Where it isn't (a sent tender, a published page), we flag it before doing it
  • Short Loom or written note explaining what changed and why
Phase 04 · Month-end summary

What got done. What's next.

Last few days of the month

End of each month, you get a written summary — what shipped, what didn't, what we'd recommend next. Same shape every time, so you can scan it in five minutes. Quarterly we do a deeper review with a strategy reset.

  • What shipped, with links
  • What didn't, and why
  • Numbers that moved (where we have them)
  • Recommendation for next month's priorities
The operator-grade reality

AI reads the mess. We hold the line.

The "too much admin" problem is three things stacked: software that doesn't fit how real jobs run, an app patchwork no single tool covers, and the stale-knowledge tax of tools used too rarely to stay fluent in. AI removes all three at once; that's the offer, and the services page breaks it down in full. What matters here is how it runs month to month:

  • AI holds the mess as it lands — email threads, WhatsApps, supplier substitutions, mid-job scope changes — and produces the document on the other side.
  • One workspace, not six apps. The contracts manager's job stops being "translate across the patchwork" because the patchwork's gone.
  • Single-approver discipline: drafts queue for the MD or their right hand. Nothing customer-facing leaves the building autonomously until trust is built.
  • Continuity month to month — context doesn't reset between cycles. The system learns the firm.
  • Same platform we run our own back office on. Operator who builds, not consultant who slides.
Why this is different

Not better than traditional consultancy. Different.

Traditional consultancy Intelligent Operations
Pricing model Day rate or fixed-fee project Monthly retainer · same fee, scope flexes
Tooling Slides, spreadsheets, Word docs, your inbox Custom workspace built per client
AI in the loop A summarising chatbot, sometimes Custom agents doing real, scoped work — the contracts-manager / estimator / document-controller layer
How a tender response gets built Sub-contracted to a bid writer · weeks · £3k–£8k Drafted from your last ten wins · days · part of the retainer
What happens after launch New scope, new fee Same retainer, ongoing iteration
Who you talk to Account manager → analyst Senior operator. One person. Always.
What you get at the end A 40-page deck A business that runs on the changes we made
The rules we run by

Six principles. Mostly learned on site.

01

Fixed retainer. No surprises.

£2k–£6k a month, rolling monthly, single quote against scope. No minimum term. If scope changes shape, we agree it in writing first.

02

One point of contact.

A senior operator. No account managers, no handoffs to analysts. The person you meet is the person doing the work.

03

Weekly check-in, monthly summary.

Enough rhythm to catch issues; not so much that meetings become the job. Quarterly for strategy resets.

04

Honest exit, any month.

Rolling monthly with 30 days' notice — if the work hasn't shifted what you do that day in a way you can see and feel, you walk. We'd rather close a misfit cleanly than carry it. No minimum term to negotiate out of.

05

Your tools, where they work.

We don't show up demanding everyone switch CRM. Keep what earns its seat; replace only what doesn't.

06

Two clients per quarter. That's the cap.

Owner-led delivery has a real ceiling. When the next slot is full, it's full. The discipline protects the depth we sell.

Starting point

Every engagement starts with a thirty-minute call. Free, no pitch.