Commercial proposals generated in minutes

The context

In digital strategy consulting for mid-sized companies, a team of 15 consultants manages 20 to 30 active opportunities simultaneously. Each requires a personalized proposal that includes client analysis, project scope, methodology, timeline, and budget.

The proposal creation process is typically entirely manual: the consultant gathers client data from HubSpot, reviews similar past proposals to reuse sections, drafts the content in Google Docs, and finally formats it into a PDF using the corporate template. The full process takes 1 to 2 days per proposal.

The challenge

The bottleneck isn't content quality — the consultants know exactly what to write. The problem is the time consumed by the mechanical process: searching for data, copying sections from past proposals, adapting language to the client's industry, and formatting the final document.

With 20-30 active opportunities, the team can only produce 3-4 proposals per week. This means many opportunities go cold while waiting for their proposal, and the team turns down potential projects simply because they don't have capacity to prepare more proposals.

The solution

The flow is built with Make and triggers when a consultant marks an opportunity as "ready for proposal" in HubSpot. The bot automatically extracts all relevant data: company, industry, size, primary contact, consultant notes, and requested services.

With that data, the system uses the OpenAI API to generate a complete proposal draft. The model is trained on the firm's 50 best historical proposals, so it replicates the tone, structure, and level of detail the team expects. The draft includes client analysis, proposed scope, methodology, estimated timeline, and investment range.

The draft is automatically formatted in a Google Doc using the corporate template and a PDF is generated ready to send. The consultant receives a notification with the document link to review and make final adjustments. On average, adjustments take 5-10 minutes. Typical implementation takes 6 days.

Results

Time per proposal drops from 1-2 days to under 10 minutes of human work. Consultants review and adjust instead of creating from scratch.

Proposal generation capacity triples: from 3-4 per week to 10-12, without adding staff. This translates directly into more closed opportunities and a healthier pipeline.

An added benefit: proposal quality becomes standardized. When all proposals start from a high baseline, the consultant only needs to personalize the nuances. The result is proposals that are consistently good, not occasionally brilliant.

Lessons learned

• Training the model on real successful proposals is the most important decision. A generic model generates correct content but without the firm's personality.
• The human review step isn't a system failure — it's a feature. Consultants add nuances that AI can't capture, and that final touch makes the difference.
• Automating proposal generation doesn't just save time — it changes the team's mindset. When creating a proposal costs 10 minutes instead of 2 days, you stop being selective and start pursuing more opportunities.