Why we built Bidwin on Claude

By Todd Heater · Published 27 July 2026, updated 30 July 2026 · Version 1.3 · An insight paper for UK bid teams

A prospect asked us recently why the whole thing is built around Claude. She'd read the site and understood the product, and her question was a fair one. You're asking me to take out a second subscription with a company I didn't choose. Tell me why that one.

The answer has less to do with model quality than she expected. It comes down to three things. Where your files end up. Whether a method can be installed or just written down. And what we didn't have to build.

Reason one: your files stay in your own folders

Think about what a bid writer handles in an ordinary week. A tender pack under embargo. An evidence library built up over years. Case studies naming real clients and real contract values, and draft responses a competitor would love to see. Where that material goes isn't a technicality.

Bidwin is a plugin. You install it into your own Claude account and point it at folders on your own machine. Claude reads and writes only in the folders you've connected, with the permissions you set. Plugins you add yourself are stored locally. Your tender pack and evidence library stay where they already sit, on your own file system and under your own backups.

So we don't host your files and never receive a copy of them. There's no Bidwin server holding your evidence library, because there's no Bidwin server. When a licence ends we have nothing of yours to return or delete.

Most AI writing tools work the other way round: the first step is to upload your material so the vendor's system can index it. That's a perfectly ordinary way to build software. It's harder to explain to a client whose confidential procurement documents you're holding, and harder still to answer when a tender asks you to describe your data handling.

Reason two: a method that installs

Most AI bid-writing advice amounts to a set of prompts. Someone good at bidding works out how to ask well, writes it down and shares it. Useful, but fragile. Prompts get half-remembered and badly adapted, and two people working from the same prompt library will produce different work without being able to say why.

Claude supports installable plugins that bundle skills, and a skill behaves the same way every time it runs. So we could build the method itself into the tool. It reads a tender pack and breaks it into scored questions with their weightings and word limits. It catches the amendment buried in the clarifications log. It drafts against your own evidence, with every figure traced to a named file. Where your evidence doesn't cover a question, it says so instead of inventing something. Then it puts the completed pack in front of an independent reviewer.

None of that is a prompt. It's a process with checks in it, developed by Mark Cosens over years of live bids. Installing it is what makes it repeatable by somebody who isn't Mark, on a Thursday, with a Friday deadline.

There's a second benefit for anyone new to AI, and it may be the bigger one. A skill carries its own instructions, so you don't need to be a prompt engineer to get professional output. Typing /bidwin-tender-vet supplies the context, the sequence, the evidence rules and the quality checks a specialist would otherwise have to know to ask for. A tool that only rewards people already fluent in AI isn't much use to a bid manager with better things to learn.

Reading the whole pack

A tender is rarely a single document, and the marks are often in the join between two of them. A criterion in the scoring annexe that a clarification quietly amended three weeks later. AI tools have a limit on how much text they can hold in view at once, so a tool working through a long pack in pieces can read both documents and still miss that the second changed the first. Claude's limit was generous enough to take a pack whole, which is why /bidwin-tender-vet reads every file in your Tender/ folder before it plans anything. Every serious model is improving here, so we expect it to matter less over time.

Reason three: we did not build our own platform

Some providers have built their own tools. The respectable reason is wanting to control the product end to end. The commercial reason gets stated less often. A platform is a moat, and once your evidence library, templates and bid history sit inside a vendor's system, leaving gets expensive.

We worked out what building our own would involve. A document editor. A file browser and a permissions model. Accounts and billing, a mobile app, a scheduler, and integrations with Google Drive, SharePoint and Slack. Then years of maintaining all of it. Every one of those already exists in the Claude desktop app, built by a company with rather more engineers than we have, and our clients are paying for it anyway.

So we didn't build it. Bidwin is the bid method alone. Here's what the app gives you that we'd otherwise have had to write, and that you get without us charging for it.

Projects and folders that keep work separate. Each tender is its own project, with its own local folder, instructions and memory. What Claude learns on one bid doesn't carry into another. Our folder convention of Tender/, Support Files/ and Instructions/ depends on the app being able to scope work to a folder and hold it there. A bid manager running four live tenders needs four sealed rooms, and the app provides them.

Draft preview with highlight editing. When Claude drafts a document it opens beside the conversation. You highlight the sentence you want changed, select Edit with Claude, and type the change. Bid managers already work this way, marking up the paragraph in front of them. It makes review closer to redlining on paper than to describing a paragraph from memory in a chat window.

Connectors. Google Drive, SharePoint, OneDrive and Slack let the evidence base stay where the organisation already keeps it. For a team, that's where the interesting workflows are. The bid manager writes while the operations lead drops the latest Ofsted report into the shared drive, and the draft picks it up.

Dispatch. You can hand a task to Claude from your phone, have it run on the desktop, and collect the finished work later. It keeps a bid moving when you're away from your desk. It's in beta, it needs the desktop app open, and it's currently on the Pro and Max plans.

Scheduled tasks. This one carries real weight for us. Bidwin Watch, our tender opportunity finder, is a scheduled task. It runs to its cadence, searches the live notices, qualifies what it finds against the client's evidence base, and reports what's worth bidding for. Without scheduling built into the platform we'd have had to run a server ourselves, which would have meant holding client data and undoing the first reason in this paper.

None of this is our work and we claim no credit for it. That's the point of setting it out. Every item is development we didn't do, time we didn't spend, and cost you don't carry. That time went into tender deconstruction, evidence discipline, house-style derivation and pre-submission review instead.

Reason four: a tool built for business use

The character of that feature list isn't accidental. Folder permissions. Approval modes that govern when Claude asks before acting. Scheduling, connectors to the systems an organisation already runs, and projects with memory scoped to the job. None of that belongs to a product designed to be entertaining. It's the groundwork for getting work finished and being able to account for how it was done, which is what we needed from a platform.

The pace of release matters as much as the direction. Anthropic ships new models and capabilities regularly, and several of the features described here didn't exist when we started. For a licence holder that means our development effort goes into the bid method while the ground beneath it improves without us. The risk cuts the other way too: a fast-moving product also changes. Features arrive in beta, and some described here are in beta now. Tracking that is part of what a Bidwin licence buys. We test new models and features and revise our guidance, rather than leaving you to follow release notes.

What we claim, and what would change our mind

Our own drafting comparisons between Claude models produced no clean hierarchy. Our guidance gives a current recommendation that could change, and we revise it as new models appear.

What we're describing is an architecture we can explain to a procurement officer without hedging. Other vendors publish their own terms, and you should read them for yourself.

The bulk of Bidwin is method, not model. The tender deconstruction, evidence rules, scoring-criteria mapping, house-style derivation and pre-submission review are ours. They encode how a bid professional works, not how any particular model behaves. If another platform offered the same three things, the method would move with us and this paper would be about that platform. Those three: files that stay in the client's own folders, a way to install a method as a versioned package, and the plumbing that spares us building an application.

Three things would prompt us to move. If the local-file architecture changed, so that using Bidwin meant handing client material to a vendor's store, the first reason here would fail. If the plugin model were withdrawn or degraded, so that skills could no longer be shipped and versioned reliably, the second and third would go with it. And if a competing platform matched all of it while being materially cheaper or more capable, we'd owe our licence holders a conversation. Finding out from somebody else that your tooling should have moved a year ago is the kind of thing that ends a working relationship, and rightly so.

The cost question underneath the question

Most people asking why Claude are really asking why they're paying two subscriptions.

Bidwin Core costs £299, paid once, and the licence is permanently yours. Claude Pro costs £18 a month and goes to Anthropic. Take the smallest case, a single tender in front of you now. Buy the licence, take one month of Claude Pro, draft the bid and stop the subscription. That's £317. A single tender written by a bid-writing firm typically costs between £3,000 and £6,000. The licence doesn't lapse when you cancel, so next time a tender lands you pay for a month and nothing else. Kept running for a year, the total is £515 for one bid professional, however many tenders you put through it.

One qualification. Claude Pro carries a weekly usage allowance as well as a per-session one, and a long pack drafted in haste can reach it and leave you waiting for the cycle to reset. Pacing the work usually settles it, and our guide How to get the most from your Claude token usage explains how. Where a deadline won't wait, Pro also lets you switch on usage credits, continuing at pay-as-you-go rates under a monthly cap you set yourself. For one busy month that's normally cheaper than moving to a larger plan you'd then pay for every month afterwards.

We could have hidden the second subscription inside a higher headline price. We could have built a platform and charged you a monthly fee to keep your bid library in it. We decided against both, for the reason that runs through this paper: your account, your files, your control, and a bill you can see.

Common questions

← Back to all resources