QUESTIONS PEOPLE ASK
Questions people ask, sorted by who is asking
The questions real people ask on real sales calls, with short answers and a link to the diagram or page that shows it. Pick your role and skip the rest.
Pick your role
Bid directors and JV leadsThe sceptic who already has CopilotIT, security and AI governanceBid writers, question owners and expertsExecutives and the leadership teamExisting customers
Every answer says where on the site the point is made. If the answer is “confirmed in writing on request”, it is because we would rather give you the exact wording than an approximate one.
01
Bid directors and JV leads
We've all got Copilot. Why do we need another AI?
Copilot is a personal productivity tool. This is the bid system for the whole team. On one question with nothing loaded they look similar; on a 20-question submission across three partners they do not. See: Three inputs, one check · Twenty people, one prompt
Who actually operates it, your people or ours?
Either. Most teams run it champion-led: one AI champion (often the bid integrator) drives the tools and supports the question owners. See: How we make you even better · Enterprise: how a pilot runs
Three companies in this JV compete against each other elsewhere. Will the others see our submissions?
No. Each partner's library is private. The JV can draw insight from it; nobody else can open or download the documents. See: Share without sharing
Our three companies will set up their libraries differently. Won't that break it?
It could, which is why every engagement starts with a library design workshop so all partners populate the same structure. See: Nine steps, four libraries
Can I see where every question is up to, at any moment?
Yes. The completeness check reads every response and shows relative completeness per question and per submission. See: Know where every question stands
Does PQQ ranking matter if we did not top it?
Less than people think. The ITT is evaluated differently, and what the client thinks of you as a team carries weight the score sheet does not show. See: How much did you spend explaining it?
02
The sceptic who already has Copilot
Where does the value actually come from, and how much better is it than a well-prompted Copilot?
From consistency, coverage and review. One prompt, run the same way on every question; the whole pack read every time; every draft marked before senior people see it. See: Three inputs, one check · Twenty people, one prompt
Does it draft like an author, or like Copilot with bullets and generic language?
It writes to your tone of voice and key messages, which are loaded once and applied to every tool. See: Three inputs, one check · Twenty people, one prompt
Our past submissions are not all brilliant. Rubbish in, rubbish out?
It wants your method and evidence, not your prose. Everything gets rewritten; what it needs is how you actually did the work. See: Same tool, different inputs
We have five years of submissions with client feedback and scores. Can it use all of that?
Yes. That archive is your knowledge core, and it transfers between programmes because the requirements do. See: Same tool, different inputs
03
IT, security and AI governance
Where does our data go?
It stays inside the Microsoft environment. Blue Edge AI runs within the Azure boundary and uses Microsoft-provisioned OpenAI and Claude models. Nothing leaves the tenancy. See: Where your documents go · Trust and data handling
Does the model train on our documents?
No. The models are provisioned through Microsoft and do not train on your documents. The exact statement is confirmed in writing on request. See: Where your documents go · Trust and data handling
Do we have to upload documents, or can it read our SharePoint?
Either. SharePoint sync, scoped to your own library, is the low-friction route. Upload works equally well. See: Where your documents go · Trust and data handling
Our parent company has strict AI governance. Will this get through?
It has been through the IT and governance review of several major European contractors. We provide the pack and join the call with your governance team. See: Where your documents go · Trust and data handling
We have Azure, a data team and Copilot Studio. Why don't we just build it ourselves?
You could build the plumbing, and the model would be the same one we use. The plumbing is not the product. What you would have to write is the question set: what the evaluator is really asking, what evidence scores, which checks run in which order, and how a draft is marked against the criteria. Then you would have to keep it sharp as the models change every quarter. That set took twenty years and £100bn of bids to write, and it is what turns a smart answer into a really smart one. See: Where the real intellectual property sits · What you choose when millions are on the table
Is this another sticky service we will not be able to get out of?
No. Your documents stay in your own document store — SharePoint, for example. Your libraries, folder structure, tone of voice and every draft are ordinary files, and nothing is held in a format only we can read. Leaving is three steps set out in writing before you start: export, delete, carry on. It sits beside Microsoft 365, Azure and the rest of your estate as one more governed service you can see, govern and leave, not a new kind of risk. See: The elephant in the room: will we be locked in?
Is it connected to the internet?
Only through clearly labelled deep research tools that you choose to run. See: Where your documents go · Trust and data handling
Which models does it use? Can we choose?
OpenAI and Claude, selected per task by Blue Edge. An extra-thinking mode routes to the larger models when you want it. See: Where your documents go · Trust and data handling
What happens to our data if we stop?
Confirmed in writing on request; we would rather give the exact statement than an approximate one. See: Where your documents go · Trust and data handling
04
Bid writers, question owners and experts
Does this replace the bid writer?
No. The writer stays in charge. The platform suggests; the team accepts, adapts or rejects, every time. See: Three gaps, not thirty pages
I own a section nobody else wanted and it spans five disciplines. Does it help with that?
Yes. That is the case it was designed for: one owner, many domains, thousands of client documents. See: Three gaps, not thirty pages
I'm the SME. Do I have to write this?
No. You close the gaps. The draft exists; you supply a crew name, a figure, a date. See: Three gaps, not thirty pages
Our reviewer is a technical specialist who wants the detail, not the sell. Does RTFQ still apply?
Yes, and the breakdown tells you which kind of reader you have. Answer what was asked, in the depth the evaluator needs. See: Three gaps, not thirty pages
It keeps telling me I've missed something. When is it enough?
The assess tool reports against the criteria. When every criterion shows its marks covered, stop. See: Three gaps, not thirty pages
05
Executives and the leadership team
What does this cost us if we do nothing?
Engineering weeks spent on content that never reaches the page, senior people fixing small errors at red team, and a bid that says different things in different rooms. See: What you choose when millions are on the table · How much did you spend explaining it?
Is this a tool purchase or a change programme?
Both, and we say so. The platform is the tool. The library workshop, the champion model and our bid team alongside yours are the change programme. See: What you choose when millions are on the table · How much did you spend explaining it?
Why Blue Edge rather than a bigger vendor?
Twenty-five years of bid consultancy built into the method, a full-time team refining it as the models change, and a business that sits alongside your bid team rather than a licence and a helpdesk. See: What you choose when millions are on the table · How much did you spend explaining it?
Why don't we just build it ourselves?
Because the expensive part is not the software. It is knowing which questions to ask of the pack, the evidence and the draft, and keeping those questions sharp as the models change. Building means a year or more before the first bid benefits, questions written by people who have not run bids, and a bid team too busy bidding to help. Blue Edge is live on one question set in weeks. The catch, stated: it is not cheap, and it is a change programme. See: What you choose when millions are on the table · Where the real intellectual property sits
Are we signing up to a sticky service?
No, and we would rather earn the renewal than write it into the exit. What stays yours: your documents, libraries, structure, tone of voice and every draft, as files. How you leave: export, delete, carry on, agreed in writing before you start. Enterprises already run on Microsoft 365, Azure, SAP and Salesforce; the question is never cloud or not, it is whether you can see, govern and leave each service. A supplier you cannot leave is a supplier you cannot trust. Stickiness is earned by the result on the next tender. See: The elephant in the room: will we be locked in?
Are you working with our competitors on this tender?
Confirmed in writing on request; we would rather give the exact statement than an approximate one. See: The elephant in the room: will we be locked in?
06
Existing customers
What's new?
Release notes go to customers first from now on; ask your Blue Edge contact for the current list. See: How we make you even better
Can you help us bring a new JV partner onto our set-up?
Yes. The library workshop is run for the partner, using your taxonomy, and their library stays private to them. See: Share without sharing · How we make you even better
Start with one conversation
Bring a live bid, or a tender you have already submitted. Thirty minutes, no preparation needed.