The Tender Qualification Agent

The requirement was on page 41.

Callum, an associate at an architecture practice in Glasgow, had spent most of three days on the quality submission for a council's primary school extension: the methodology, the team CVs, two case studies rewritten to match the scoring guidance. On the Thursday before the deadline, Fiona, one of the four partners, sat down to sign off the Single Procurement Document and read the insurance schedule properly for the first time. The contract notice had only pointed to the tender documents.

Professional indemnity cover of £5 million, each and every claim, evidenced at submission. The practice carries £2 million. The broker could quote an uplift, at a price the fee would not carry, and not by Monday.

They withdrew. Thirty-one hours of work went into a folder marked "not submitted", and the requirement that killed it had been in the tender pack since the day it was published.

The practice, Fiona and Callum are a composite, built from the patterns we see in practices of this size, and the figures attached to them are design projections. The regulations, the portal and the published figures are real, and linked to their sources.

A 24-person architecture practice in Glasgow: four partners, mostly education, healthcare and community buildings, about £2.1 million a year in fees, a little over half of it from public clients. Public work in Scotland is advertised on Public Contracts Scotland, which Scottish public bodies must use for regulated procurements, £50,000 and over for services. The practice gets the portal's email alerts for architectural services, and a partner or Callum reads every notice that looks close.

In the last twelve months that came to about 140 tender packs worth opening properly. The practice bid for 38 and won 7.

This Blueprint is about the decision before the bid: which tenders to write at all. Writing the bid itself is a different problem, which The Proposal Agent covers.

Nobody in the practice is careless about bids. The partners are careful, the bid writing is good, and the team knows which clients it does well with. The trouble is the order in which things happen.

The reading comes before the decision. A tender pack is a contract notice, the Single Procurement Document, a specification, scoring guidance and a draft appointment, often a hundred pages or more. Nobody decides whether to bid until somebody has read enough of it to feel sure, and reading enough takes about four hours. So the go/no-go decision is made after the most expensive part of qualifying, by whoever happened to have time that week.

The knockouts are scattered and specific. Most tenders the practice should not bid for can be ruled out on a handful of facts: the turnover required, the insurance required, the experience required and how recent it must be, and how heavily price is weighted. In Scotland those requirements are meant to be proportionate. For contracts above the procurement thresholds (for a council's services contract, £207,720 including VAT from January 2026), the regulations say a minimum turnover requirement must not exceed twice the estimated contract value, except in duly justified cases, and the buyer must then state its reasons. Between £50,000 and those thresholds, the buyer still has a general duty to act in a transparent and proportionate manner. But each fact sits in a different document, in different words, and the one that matters is easy to miss until somebody signs the declaration.

Price weighting is read, then ignored. Above the same thresholds, Scottish rules do not allow an award on price alone, and the Scottish Government's construction procurement handbook encourages buyers to promote quality. Some buyers still weight price at 60%, often enough for the Royal Incorporation of Architects in Scotland to say in July 2025 that 40% quality against 60% price sends the signal that safety and sustainability are secondary. The practice knows its record on price-heavy tenders is poor, and bids on them anyway, because by the time anyone has read far enough to notice the weighting, two people are already invested.

Run the past year's 38 bids back through a simple set of knockout and fit rules, and we model 14 of them as no-bids from the start: three failed an insurance or turnover requirement, four lacked recent experience of that building type, and seven weighted price at 60% or more. The practice won one of those 14, at a fee the partners later called a loss. It won six of the other 24. We set the rules with last year's results in view, so treat 14 as the most the rules could have caught.

At an average of about 55 hours a bid, the 14 no-bids cost roughly 770 hours. Reading the 140 tender packs took about 560 more. Together that is around 1,300 hours a year, about three-quarters of a full-time architect's working year, most of it spent reading and writing for work the practice could not win or did not want.

Public bidding is a low-odds game. An independent review of construction frameworks for the UK Government put a supplier's combined chance of getting onto a framework and then winning a project from it at 20 to 25%. Against that, the practice's 18% on public tenders is low, though not by as much as it sounds. What it controls is which tenders it sits down to write.

The design is next. The design, the build guide, the running costs and the take-away are free to subscribers. Enter your email below and the rest of this Blueprint opens. New ones arrive every Tuesday and Friday.