Tender drafting and pricing analysis grounded in company evidence
The work connected a company's evidence to automated tender drafting and used historical quality-score analysis to inform commercial pricing decisions.
- Scope
- Company knowledge, tender workflow and pricing analysis
- Work status
- Previous implementation
- Relevant offering
- Managed business functions ↗
Explain how you will keep the client informed.
Approved communication procedure
Source S1 · sample materialA named lead coordinates the agreed reporting schedule.[S1]
Source reference stays with the wording.Bringing evidence and commercial judgement together
A quality-led tender requires both a credible response and a clear understanding of how the buyer evaluates quality and price. The work connected three questions: what the company could substantiate, how to use that evidence in a response, and how quality scoring should inform pricing.
A company-evidence library for tender drafting
The work created a searchable company-knowledge system linked to CRM data and resource sheets. An automated tender-drafting workflow used that source-backed information to support responses with traceable provenance.
The underlying approach used GraphRAG to connect related information. Its purpose was to give the drafting process a structured foundation of company facts and evidence. Traceable sources support checking; they do not make an AI-generated response automatically error-free.
Analysing quality scores to inform price
James reports analysing 750 historical tender results to understand quality scoring. A quality-headroom algorithm then examined how much additional price a quality advantage could support under the relevant evaluation rules.
The analysis informed a pricing decision rather than removing commercial judgement. Its interpretation depends on the scoring formula and the information available about competing bids.
Reviewing whether a discount changed the award
The work also examined final-sign-off price reductions. James reports that, across the bids reviewed, reconstructing the scores at the original prices left the award outcomes unchanged: the lost bids would still have lost, while the winning bids would still have won.
That was a retrospective calculation for a particular reviewed set. The reconstructed results are distinct from the actual awards and do not show that discounting is unnecessary in every tender.
How the work connects
From the information.
To the decision.
- 01Input
Identify the requirement
Start with what the buyer asks for and the quality and price evaluation rules.
- 02Workflow step
Connect the evidence
The company-knowledge system links relevant facts with CRM data and resource sheets, retaining provenance.
- 03Human input / decision
Prepare the response
Source-backed information supports the automated drafting workflow and gives reviewers evidence to check.
- 04Human input / decision
Inform the price
Quality-score analysis and previous results support a commercial decision under the relevant scoring rules.
Quality and price.
Read together.
Quality pointsPrice points
In this example, the quality advantage more than offsets a higher price.
Assumptions: quality carries 60 points and price 40. Raw quality scores are 90/100 for Bid A and 80/100 for Bid B. Bid B has the lowest indexed price, 100.
Formula: quality points = raw quality × 0.6; price points = 40 × lowest price ÷ bid price. Total = quality points + price points.
Prices are synthetic indices, not currency. Different scores, prices or evaluation rules can change the result.
The result
What changed
The drafting process gained a traceable company-evidence foundation, and pricing discussions gained analysis of quality scores and previous results. James reports that the approach was used during a period of improved tender performance, with more wins at higher margin.