How To Use Your Response Library To Tender With AI

TLDR: Don’t upload whole tender documents into AI and expect great outputs. Instead, use your Response Library and guide the AI with your expertise and involvement. Working systematically answer-by-answer will result in better responses in less time.

The best way to start using AI is with human involvement.

The more specific you can be about what you require AI to write, the better outputs you’ll get. 

And while it requires a little bit more effort upfront, you’ll spend less time editing if you can give a really clear set of instructions to work with.

Plus, you’ll get a feeling for how the models behave, what works well, what doesn’t and how you like to use it in your workflow.

The bottom line: Your response library + AI + your guidance = better submission outputs.

Here’s How You Can Use Your Response Library For AI Success

Tip #1: Work Question by Question

You’ll get better writing from AI if you can be really specific, giving the AI the specific context required for an answer.

Don’t upload a 50-page tender document and ask “write me a response”.

Instead, extract individual questions and work systematically:

• Upload 2-3 (or more) relevant past responses from your library
• Include the specific new question
• Add specific subsections of the tender documents to add breadth to the AI’s context (adding evaluation criteria is a great way to guide the AI)
• Add any additional context about your business, such as product spec sheets, policies or documentation
• Then request the response from AI

Getting your AI context together ahead of time requires a bit of extra investment, but the improved responses will mean you spend less time editing.

Tip #2: Structure First, Content Second

For longer, more complex responses, always scaffold the structure before asking for content. Don't just leave the AI to decide this (or you'll end up with lots of editing later).

Provide your response structure first, then ask for content within each section, again using the relevant responses from your Response Library.

What I find works well is to write dot points for what you want to say with a response, and then get the AI to fill in the rest. This way, you can focus on "what will be an impactful statement" rather than worrying about exactly how to word it.

Tip #3: Use Separate “Conversations” for Separate Responses

Regardless of which AI tool you’re using, it’s probably got a chat interface.

The chat interface can be a great way to interact with AI models, but unfortunately it also masks one of the biggest limitations with AI models - context length.

AI has a limited “context” - this is essentially its memory of your current conversation. It can be easy to keep working in the one conversation when you're working on a tender, but you will get much better results by having shorter, more focused conversations.

Every time you start work on a new question, start a new conversation and just include the relevant information for that answer in the chat.

Summary

When getting AI to write your responses:

1. Use your response library and give the AI only the relevant information for the question you are answering.
2. Scaffold longer responses first - write the broad structure first (including the key sections, points or statements) and get the AI to fill in the rest based on your response library.
3. Always use “new conversation” when you’re starting on a new question.