The Snagging List That Refused to Die: An Anatomy of Open Defects on a Reading Office Fit-Out

Every commercial mechanical contractor has a snagging list they'd rather not discuss. Ours, mercifully, got closed. But we recently reviewed the paperwork from a Reading office fit-out we inherited mid-project — a 4,000 square metre CAT B fit-out with an AHU replacement, LTHW modifications and a full BMS reconfiguration — and the snagging file told a story worth telling, because it is the same story playing out on sites across the UK right now.

The list began, as they all do, with 38 items. Eighteen months later it contained 214. Nothing had been added by way of new works. The list had simply grown.

How a List Grows Without Any Work Being Done

Snagging lists don't expand because buildings develop new defects. They expand because the definition of a snag drifts. Item 4 — 'AHU 2 door latch stiff' — is a snag. By month six, the list includes 'BMS graphics to match client brand guidelines' and 'confirm F-gas log format acceptable to end user's FM team'. These are not defects. They are negotiations wearing a snag's clothing.

The mechanism is simple and almost always the same. The main contractor's commercial team realises that an open snag list is the only leverage remaining once retention is the last card on the table. The client's FM team realises that adding items delays the point at which the contractor stops answering the phone. Both parties feed the list. Neither intends to close it.

On the Reading project, the original 38 items took eleven weeks to resolve. The remaining 176 items — most raised after practical completion, during the 'defects liability period' that sounds contractual but behaves like a suggestion — were still open when the building's second maintenance regime was already in place.

Who Actually Benefits from an Open List

It's tempting to blame contractors, and sometimes the blame is deserved: the flue detail was wrong, the condensate falls were wrong, and the insulation around the buffer vessel was an act of interpretive sculpture. But the incentives run in several directions at once.

The main contractor benefits because an open list defers final account. The client's procurement team benefits because an unclosed project is harder to audit. The incoming FM provider benefits because every open item is a reason the previous party still owes attention. And the mechanical contractor — the party with the least leverage — benefits least of all, because they're holding labour on site for items that were never properly specified in the first place.

The result is a strange equilibrium in which nobody wants the list closed, and the only party who wants it closed is the one who can't force it.

The Five Per Cent That Lasts Five Years

Ask any building services engineer which snags actually get done and you'll hear the same pattern. Items affecting life safety, comfort or compliance — fire dampers, TMVs, emergency lighting interfaces — get closed quickly, because the law is impatient. Items affecting nobody's KPI — a mislabelled valve, a motorised valve that hums, a weather compensation curve set two degrees conservative — linger indefinitely.

On the Reading job, the last item closed was a BMS sensor calibration on a riser serving a floor that had been sublet to a tenant who never reported a problem. The sensor had been reading three degrees out for two years. The energy bill absorbed it. Nobody's dashboard flagged it. It was only closed because we found it during a separate maintenance survey.

That is the real cost of the unclosed list: not the paperwork, but the defects that stop being defects simply because everyone stopped looking.

Contract Language That Actually Forces Closure

We're not contract lawyers, but we've sat through enough practical completion meetings to know what works. If you're a client, a landlord or a main contractor procuring mechanical works, three clauses do more than any amount of chasing.

First, define a snag. A snagging item is a defect against the specification or drawing — not a preference, not an enhancement, not a documentation request. Anything else goes on a separate list with a separate commercial treatment. This single distinction prevents most list inflation.

Second, cap the defects list at practical completion. State in the contract that the defects liability schedule is fixed at PC, that items added afterwards are dealt with under the variation mechanism, and that the list must reduce to zero — not 'substantially zero', not 'save for matters of no material consequence' — before final certificate and retention release.

Third, attach a clock to every item. Each defect carries a remedy period. If the contractor fails to remedy within it, the client may engage others and recover the cost. This exists in most JCT and NEC frameworks already — but only bites if somebody keeps the schedule current. Appoint one named person to own the list. Not a team. A person.

The Uncomfortable Conclusion

Snagging lists never close because closing them requires someone to stop benefiting from them being open. The fix isn't goodwill or better project management software. It's contract language that makes closure the cheapest option for every party — and a named individual with the authority to say, in writing, that the list is done.

On the Reading fit-out, closure came two years late and cost the client a recalibration survey nobody had budgeted for. The building is fine now. The list, we're told, is archived rather than closed — which is perhaps the most honest description of how these things actually end.