BidWriter is a KnowledgeFlow assistant that helps your team write better tender responses, grounded in your organisation’s own documented experience. It reads the tender pack, finds every question, works out where the marks are, and drafts evaluator-focused answers using evidence retrieved from your indexed library of past bids, case studies and project records.
This guide gets you from a standing start to a scored, drafted response. Allow 10 minutes for your first run, or 45 minutes to work through a full bid.
Before you start
You will get far more out of BidWriter if you have these to hand:
- The tender pack — the ITT, RFP, RFQ or PQQ document itself.
- The supporting documents — specification, background and client context, annexes.
- The evaluation criteria and weightings. This is the one that matters most. BidWriter uses the weightings to decide where to concentrate evidence and effort, so without them it cannot tell a 40-mark question from a 5-mark one.
You can upload files or paste text straight into the chat. Both work equally well.
Three things to know about how it works
These are deliberate design choices. Knowing them up front will save you a lot of second-guessing.
It pauses for your confirmation. The question-by-question workflow completes one step, checks with you, and only then moves on. It never runs ahead. This keeps you in control of a live bid — it is not the tool being slow.
It shows its working while you draft. Draft answers carry inline citations to the case studies and past bids each claim came from, plus visible gap flags where your indexed evidence does not support something the question needs. These are there so you can verify and fill gaps. The final assembled response strips them all out.
It will not invent experience. Every claim about your organisation comes from your indexed library. A fabricated claim in a live bid is worse than an honest gap, so gaps are flagged rather than papered over.
Your first 10 minutes
The quickest way to see what BidWriter does is to decode a live tender.
- Open BidWriter and press What is the RFP asking for.
- Share the RFP or ITT — a file or pasted text.
- Read what comes back.
You should get a plain-English summary of the core ask, a list of what bidders must actually deliver with the procurement language stripped out, any red flags (tight mobilisation windows, named-individual requirements, low price weighting), and any contradictions between the specification and the scoring.
You will know it worked when you can explain the opportunity to a colleague in two sentences without opening the tender pack.
The four workflows
Each button starts a different job. Pick the one that matches where you are in the bid.
What is the RFP asking for
Use it when: a tender lands and you need to understand it fast, before committing anyone’s time. You get: the real ask in plain English, the substantive requirements one per line, red flags and contradictions called out separately, and an honest view of what it takes to deliver the contract — not just to win it on paper. Note: this one works only from the document you share. It does not draw on your indexed library, and it does not make a bid/no-bid call.
Respond question-by-question
Use it when: you have decided to bid and are ready to write. You get: a supervised, step-by-step drafting process. Fresh evidence is retrieved for every single question, so each answer surfaces the closest-matching case studies rather than recycling one generic set. This is the core of BidWriter and the workflow most teams spend their time in. See the walkthrough below.
Score my response
Use it when: you have a draft — yours or a past submission — and want to know how an evaluator will mark it. You get: the tender’s own marking scheme extracted and shown back, a scorecard per question with a realistic mark and justification against the scoring descriptors, compliance and word-count flags, an evidence check that names stronger proof sitting unused in your library, and a prioritised list of fixes ranked by marks at stake. Note: this is an evaluator’s review, not bid copy. It is deliberately blunt.
Create Case Study
Use it when: you have finished a project and want it working for you in future bids. You get: a fact sheet drawn from your project record for you to confirm, then a client-ready case study in your house format, carrying the metadata and tags that let future tender responses find and reuse it. Note: this is the one workflow where your uploaded documents are treated as the factual source, because you are uploading the record of your own delivery. Quotes are reproduced exactly or omitted and flagged — never invented or improved.
Walkthrough: writing a response end to end
- Start. Press Respond question-by-question and share the full tender pack, including criteria and weightings. BidWriter runs a document check first and will stop and ask if something important is missing.
- Confirm the question list and scoring map. It extracts every question — including sub-questions and requirements buried in the specification — and maps the marks, thresholds and pass/fail conditions. Check this against the tender yourself and correct anything it has missed. Everything downstream depends on it.
- Draft one question at a time. Each answer opens with the customer benefit, builds through delivery and cited evidence, and closes on compliance and value. Read the citations. Ask for changes; it will redraft.
- Fill the gaps. Where a question needs something your library does not evidence, you will see a
[⚠ GAP]flag telling you exactly what is missing. Supply the detail, or write that portion yourself. - Run the review. Once every question is drafted, ask for the compliance review: completeness, criteria coverage, effort against weighting, word counts, outstanding gaps, and repetition across answers — including the same case study appearing three times.
- Assemble. Ask it to assemble the final response. You get one continuous block of clean bid copy with every citation, flag and internal note removed, ready to export to Word from KnowledgeFlow.
Getting better results
- Use a real tender your team knows well, especially the first time. You can only judge the output if you can judge the evidence behind it.
- Always upload the evaluation criteria. Effort follows the marks, and BidWriter cannot follow marks it cannot see.
- Correct the scoring map before drafting. A missed sub-question at step 2 becomes a missed answer at submission.
- Read the citations, not just the prose. The fastest quality check is asking whether these are the projects you would have chosen.
- When an answer reads thin, say which project it should have used. That tells us whether the gap is in your indexed library or in the assistant’s instructions — and it is the single most useful piece of feedback you can give us.
- Push back. Ask for a different case study, a shorter answer, a sharper opening. It redrafts.
What it will not do
- Invent experience, metrics, clients, awards or accreditations. If it is not in your library, you get a flagged gap instead.
- Present your client’s own specification or background as your organisation’s track record.
- Make the bid/no-bid decision for you.
- Guarantee a score. The marks it predicts are an informed estimate — it cannot see competing bids or a panel’s discretion.
- Export the Word document itself. BidWriter produces the text; KnowledgeFlow handles the export.
Common questions
It has stopped and asked me to confirm something. Is it stuck? No. The question-by-question workflow is supervised by design. Confirm or correct, and it will offer the next step.
Why are there citations and warning flags all over my draft? So you can verify the evidence and see the gaps while there is still time to fill them. They are stripped from the final assembled response.
Can I paste text instead of uploading files? Yes, for every workflow. Partial or informal content is fine — BidWriter will tell you what is missing.
It missed a case study I know we have. Tell us which one. That usually points to a gap in the indexed library rather than the assistant, and it is a quick fix.
Next steps
Run the five-test handbook with your team to check BidWriter against your own track record before you use it on a live submission, and send us your feedback. We use it to tune the assistant’s instructions and review your indexed library with you.