Retention is the quiet margin killer on shopfitting contracts. Here is what to track, why spreadsheets struggle, and what changed in 2025.
On a ten-contract shopfitting ledger, retention is rarely the thing that keeps an FD awake at night. A dispute with a main contractor, a cash-flow wobble, a slow-paying customer — those are what make it onto the weekly call. Retention sits quietly in the background: 3–5% off the top of every application, held for twelve months after practical completion, paid back, usually, after the defects liability period ends.
Until it is not. A main contractor insolvency in month eight. A retention release “under review” for fourteen months. A ten-year-old retention balance nobody can find the paperwork for. Add them up and retention is the difference between a profitable year and a merely busy one.
This article is for Finance Directors, Credit Controllers and MDs who want a practical, 2026-ready view of retention tracking in construction. It covers the regulatory backdrop, why spreadsheets fail, what a healthy system looks like, and where Contract Controller fits.
Retention is the percentage of each application for payment that your customer holds back — typically 3% or 5%, occasionally 10% on public-sector work. Half is usually released on practical completion. The second half — known as the second moiety — is released at the end of the defects liability period, usually twelve months later.
On a £1m contract with 5% retention, you are looking at £50k held over the project, £25k of which is locked for a year after you have walked off site. Across a live order book of twelve contracts, you can easily have £250k–£500k in retention outstanding at any moment.
The Reporting on Payment Practices and Performance (Amendment) Regulations 2025 added retention-specific disclosures to the payment-practices reporting regime. From March 2025, qualifying firms — broadly, UK businesses above the reporting threshold — must report on retention amounts withheld and released as part of their bi-annual filings. Commentary from FIS and the NAS has been unambiguous: this is the first time retention practices become publicly visible by contractor.
Three practical implications for shopfitters:
Retention spreadsheets start simple. By contract twenty, they rarely are. Five recurring failure modes:
Usually a credit controller. One person knows the file, the assumptions, the workarounds. When they take two weeks off, retention release dates slip.
PC is negotiated, sometimes disputed, often not formally certified until weeks after site-finish. A retention schedule built on expected-PC dates will routinely chase releases that are not yet due — and miss releases that are.
The second-moiety release is almost always twelve months after PC. Spreadsheets do not recalculate the release date when the PC date changes. The result: a second moiety that sits on the ledger unchased for fourteen months, not twelve.
Variation order £30k, retention 5% — that is £1,500 of additional retention that must attach to the original contract’s release schedule. If the variation is logged in a separate workbook, the retention figure is wrong from day one.
If a main contractor disputes whether you chased retention in good time, a spreadsheet with no versioning gives you nothing. Dispute resolution hangs on evidence of communication.
None of these is a capability failure of Excel. They are all consequences of managing a live, contractually-governed position in a tool designed for static analysis.
Six characteristics to design for — whether you buy, build or spreadsheet your way there:
If your current system struggles on three or more of these, it is already costing you money — you just have not measured how much.
Retention in Contract Controller is a first-class concept, not a spreadsheet bolt-on. Every application for payment books its retention automatically; first and second moiety are dated and chased independently; variations carry their retention through to release without re-entry. Aged retention lives in its own view, reported separately from the main sales ledger. The practical effect is that retention stops being a thing someone has to remember and starts being a thing the system remembers for them.
For firms above the payment-practices reporting threshold, the platform also generates the retention figures required for the bi-annual filing in a single export — which, post-2025, is the difference between a one-hour reporting exercise and a three-day one.
book a short demo of how Contract Controller handles retention end-to-end at https://contractcontroller.co.uk/request-demo/
