Skip to content
PlotUProperty intelligence · AI · evidence

Procurement · 8 min read

Three quotes, one shape: a procurement workflow that finds the gap before site does

The lowest number was winning packages and the scope difference was arriving as variations. The workflow made the difference visible before award.

Published

About this case study

This is an anonymised composite drawn from recurring engagements. No client is identified, no figures are invented, and no outcome is guaranteed. The workflow is the substance.

The situation

A construction SME was letting trade packages on the basis of quotes that could not honestly be compared. Each subcontractor priced to their own read of the drawings, with their own exclusions and assumptions, and the comparison was happening — at best — in a spreadsheet built differently each time.

The details that follow are a composite of recurring engagements of this kind. No client is identified, and nothing here is a claim about a specific company's results.

How the workflow runs

  1. Force every quote into the same structure

    Each quote is extracted into a fixed format: scope items, inclusions, explicit exclusions, provisional sums, assumptions, programme and payment terms. AI does the extraction; a person checks the output against the original document before anything is compared.

  2. Build the comparison matrix

    With all tenderers in one structure, the workflow produces a line-by-line scope matrix: what one priced, another excluded, what one assumed, another queried. Provisional sums are flagged with what triggers them.

  3. Send questions back before award

    The comparison generates a clarification list per tenderer. Awards are made on resolved scope, not on the hope that exclusions will not matter.

  4. Evidence the award decision

    The award is recorded against the final comparison and the drawing revision it was priced on. When a variation is later claimed, the difference between priced scope and claimed scope is a document lookup rather than an argument from memory.

What changed

The workflow did not make procurement cheaper by magic. It made it honest: scope gaps surfaced while they were still questions, not after they had become variations.

  • Quotes are compared on scope, not on the skill of each subcontractor's PDF formatting
  • Clarification questions go out before award as standard
  • The award rationale is written down against a drawing revision
  • Variation discussions start from a shared record instead of recollection

Where the professional line sits

  • Measurement, valuation and contractual position remain with the quantity surveyor — the workflow hands them structured comparisons, not conclusions
  • Every extraction is verified against the source quote before comparison
  • No award is made on AI output alone; the comparison informs, the commercial lead decides

FAQ

Common questions

Can AI read subcontractor quotes reliably?
Text-based quotes, yes, with a human verification step. Scanned documents and drawings need more care, and no figure is relied on until checked against the original.
Does this replace a quantity surveyor?
No. It removes transcription and reformatting so the QS spends their time on measurement and judgement. Anything contractual stays with the professional.
Is this based on a real contractor?
It is an anonymised composite of recurring engagements. We publish workflows, not client identities or invented savings figures.

Keep reading

More case studies

Discuss your workflowAll case studies