The Tenders You Don't Win Are Still Data

Estimating is a volume business. Tenders arrive constantly, a fraction get qualified, fewer get submitted, and fewer again come back as work. Every commercial plumbing contractor runs at those odds. It's the cost of being in the market.

Which means the risk in an estimating department isn't losing a tender. It's losing one quietly. A tender that nobody followed up, that sat past its due date while everyone assumed someone else had it, that went cold and was never formally declined. Nothing in the numbers tells you that happened, because a tender you forgot and a tender you lost look identical at the end of the year.

Impact's director wanted the estimating team held to something. Not a win rate, which depends as much on who else bid, but turnaround and follow-up, which are entirely within the team's control. Nobody could see either. There was no record anywhere of when a tender was last actioned.

The annual view had a separate problem. Live tender reporting counts what sits in each bucket right now, which is exactly what a weekly meeting needs. It's the wrong shape for a year in review. As tenders move from qualified to submitted to won, the counts behind every earlier month move with them, so a financial year never settles. The team sat down to read a full year together and the top-line figures wouldn't reconcile against each other. The numbers weren't wrong. They just described today, not the year.

We stopped trying to make one report do two jobs.

The live report stays live, and still drives the weekly estimating meeting. Alongside it sits a locked version, frozen per financial year, where the numbers never move again. A year selector switches between the current year, last year and the one before, so a comparison takes a click instead of an argument about date ranges. The decision was deliberately unclever: half a day once a year to cut and lock the static copy, rather than engineering one report that tries to be both and fails at each.

The report leads with three labelled figures. Target, win rate, win rate required. Those are the numbers the business steers by, so they sit across the top where the weekly meeting reads them first, legible in a glance. The reports have been built alongside the business for years and keep being refined around what it needs to focus on next; this pass sharpened that further.

Each month the report is rendered to PDF and emailed out on the first business day after month end. That matters more than it sounds. A link points at a report that will have changed by the time anyone opens it, while a PDF is a fixed record of where the pipeline stood that morning, and a folder of them is a year you can compare against.

On the tender list itself, the status cards were replaced with the numbers the department actually runs on: how many tenders are outstanding, how many are overdue, and how many came in over the last seven and thirty days. Volume in, work in progress, and what's late.

Every tender now carries a last contacted date, stamped automatically with the author whenever a note is logged against it. It isn't a field anyone types into, because a field people maintain by hand is a field that lies. It's a real column written at the moment the note is created, which means the list can be sorted by it. Sort descending and the tenders nobody has touched in a month rise to the top.

Around that sit the smaller things that make a list usable: tender ID and revision number as visible columns, and an export of submitted tender notes to Excel for anyone who wants to work through them offline.

Two reports cover what the department gives away. A binning report breaks down disqualified tenders by sector and by client, so the business can see what it declines as a pattern rather than one decision at a time. A labour utilisation report shows where estimator hours are going, filtered to any single month across a two year window, with buttons for last week, last month, last quarter and the year to date.

The weekly estimating meeting reads off the report from top to bottom. The figures at the top are legible, so the meeting starts with the position rather than with someone interpreting it.

Cold tenders surface by themselves. Sorting by last contacted turns "has anyone chased this?" from a question asked in a meeting into a column read before it.

The financial year holds still. A locked year can be quoted in a board pack, sent to an accountant, or compared against the year before, and it will say the same thing in six months as it does today.

Each month adds to a record instead of overwriting one. The monthly PDF means the company accumulates a series of fixed positions, so movement between months is measurable rather than remembered.

Estimating effort became visible. Hours against tenders, by month and by sector, alongside the binning report, means the cost of chasing a sector the company rarely wins is a number now.