From Checking to Assurance: A Maturity Model for Review Teams

Most review teams are one missed comment away from a problem they never saw coming: not because the reviewers are careless, but because the way most firms organise review work makes comprehensive, confident, defensible assurance structurally impossible, regardless of how good the individuals are.

Why the industry needs a maturity language for review.

The industry has spent years talking about quality. What it hasn’t built is a language for maturity in how review work is actually done. This framework does that. Five stages. A clear line of progression. And an honest reckoning with where most firms sit today.

Stage 0: ad-hoc checking.

This is where review lives when no one has decided it needs to be designed.

A senior person looks over a document. They catch what they catch. Nothing is recorded beyond a few emails or a mark-up PDF. The next reviewer starts from scratch: no memory of what was looked at before, no record of what was resolved, no visibility into what wasn’t checked at all.

The output is a signature, not assurance. The word “checked” appears on the sheet, but it means someone looked at this. It says nothing about coverage, confidence, or what happens next time.

Stage 1: standardised checklists.

The first instinct when firms try to fix ad-hoc checking is to write a checklist. That’s not wrong. Checklists force consistency. They make review legible. They give less-experienced reviewers a scaffold. When used well, as Atul Gawande’s work on checklists in high-stakes fields has shown, they reduce preventable error.

But a checklist is a floor, not a ceiling. It confirms that the items on the list were considered. It says nothing about whether the list is complete, whether the reviewer had the confidence to tick a box they weren’t sure about, or whether the item was actually resolved or just visited.

Checklist-based review also ages poorly. Requirements change, schemes grow in complexity, and the checklist stays static: until someone realises it’s missing something important, usually after the fact.

Stage 2: full-coverage review.

Stage 2 is the first level where review starts to function as assurance rather than sampling.

Full-coverage means every requirement that applies to a scheme is identified, tracked, and assigned a status. Not just the items on a pre-written list: the actual requirements, derived from the brief, the applicable regulations, and the firm’s own standards. Coverage is measured, not assumed.

This is a meaningful step forward. It makes gaps visible before they become problems. It makes the review record useful to others on the project, not just the person who ran it. It creates an audit trail.

The limitation at Stage 2 is that coverage alone doesn’t distinguish between confident and uncertain findings. A requirement can be marked as reviewed whether the reviewer was certain or guessing. The record doesn’t carry that signal.

Under the Building Safety Act 2022 and the Gateway 2 regime for higher-risk buildings, this distinction is no longer academic. Dutyholders signing off at Gateway 2 need to know not just what was reviewed, but how well.

Stage 3: calibrated confidence.

This is the stage where review becomes genuinely decision-useful.

Calibrated confidence means the review record doesn’t just show what was looked at: it shows how sure the reviewer was, and why. A finding isn’t binary (pass/fail). It carries a grade: the reviewer’s actual confidence in that finding, backed by the evidence that supports it.

Why does this matter? Because the two things that most often cause safety incidents are not the things anyone was confident were wrong. They’re the things no one looked at, and the things the team was confidently wrong about. A graded record surfaces both: blind spots as honest “not yet checked” entries, and shaky findings as explicitly flagged low-confidence items.

An honest “not yet confident” is safer than a false all-clear.

Dame Judith Hackitt’s 2018 Independent Review of Building Regulations and Fire Safety didn’t just call for more checking: it called for a culture change, from treating compliance as a box-ticking exercise to genuinely owning the outcome. Calibrated confidence is the operational version of that shift. Stage 3 review makes both blind spots and uncertain findings visible, rather than burying them under a signature.

Stage 4: compounding record.

The jump from Stage 3 to Stage 4 is the one most firms don’t even know they’re missing.

Stage 3 gives you great assurance on this project. Stage 4 turns every completed review into an asset for the next one.

A compounding record means that every resolved judgment, every time a reviewer concluded how a requirement applies, what confidence they had, and why, is retained as firm-specific knowledge. Not locked in a PDF. Not in someone’s head. Structured, searchable, and available to the next reviewer on the next scheme.

This is where durable competitive advantage in review quality is actually built: not in the tool, not in the checklist, but in the accumulated, firm-specific record of how this organisation applies its standards, and the fact that it gets stronger with every review.

It also changes the economics of onboarding and succession. When a senior reviewer leaves, Stage 0–3 firms lose their judgment. Stage 4 firms retain it.

Where most firms sit today.

Honestly, most sit somewhere between Stage 0 and Stage 1. Not because the people are bad: the reviewers in most mid-tier AEC firms are experienced, diligent professionals. But review as a designed process has been chronically under-invested in. The tools available have been generic. The regulatory pressure, until recently, was manageable enough that informal practice held.

That’s changing. Post-Grenfell regulation, the Gateway regime, and the formal dutyholder responsibilities introduced by the Building Safety Act have moved the bar. What was once acceptable informality is now documentable liability.

Firms at Stage 0–1 aren’t just leaving quality to chance. Under the current regulatory environment, they’re accumulating exposure. Stages 2–4 aren’t utopian: they’re achievable, with the right infrastructure, the right process design, and the willingness to treat review as an engineering discipline rather than a final check.

Assurance engineering is the discipline that gets you there.

At Tektome, we call this Assurance Engineering: treating quality as a volume you raise and calibrate, not a verdict you stamp at the gate.

Tektome is built for review teams who are ready to move past checklists and build a review function that compounds over time. Full-coverage tracking, calibrated confidence, and a firm-specific record of resolved judgment, structured into the workflow, not bolted on after the fact.

The firms building this capability now will have a structural advantage that can’t be copied. The record of judgment can’t be bought off a shelf. It has to be built: scheme by scheme, review by review.