Definition
The term matters because of what it is not. It is not a chatbot. It is not proposal software. It is not a marketing agency, and it does not replace the firm’s people. Those things exist, some of them are good, and none of them holds the thing that actually wins complex work: the firm’s own judgement.
In firms that sell complex services, engineering consultancies are the clearest case, the ability to win work is concentrated in a few senior heads. They know which opportunities are worth pursuing and which are dressed-up losses. They know what the client is actually buying, what the incumbent will do, what the real objection is behind the stated one, and what a winning submission emphasises. When those people are in the room, the firm wins. When they are stretched thin, the firm loses work it should have won.
Most firms treat that as a staffing problem and hire more business development. A conversion brain treats it as an engineering problem: the judgement exists, it is simply trapped. Extract it, encode it, and run it on every deal, with the senior people governing outcomes instead of processing work.
Four components, in plain words:
Red Lines. What the system must never do. The firm’s ethics, brand standards and pricing floors, written down and enforced. The system operates inside them without exception.
Winning Rules. How the firm actually wins, taken out of its best people’s heads and made repeatable: how to qualify an enquiry, when to walk away, what to lead with for this kind of buyer, what a strong submission looks like.
Deal Memory. Every enquiry, quote, objection, win and loss the firm runs through the system, held as institutional memory. The system answers from the firm’s own record, not from the internet, and every deal it sees makes it sharper.
The Trust Dial. Every decision the system makes carries a confidence score. High confidence, it acts. Low confidence, it hands to a nominated human, and every handover teaches it. Autonomy is earned the way a driver’s licence is earned, in stages, with demotion if performance slips.
All of it. The Red Lines, the Winning Rules, the Deal Memory and the workflows live inside the business and stay there. The AI models underneath are rented muscle: swappable, vendor-agnostic, and cheaper every quarter. The part the firm owns is the part that compounds. A competitor who starts two years later cannot buy that accumulated judgement off a shelf, because it is not on a shelf.
This is the line that separates a conversion brain from the software category next door. When a firm’s knowledge lives inside a vendor’s platform, the firm is renting access to its own expertise. A conversion brain inverts that: the models are the replaceable part, the firm’s judgement is the asset.
Not AI proposal software. Tools that draft tender responses are useful at the document stage. But the document is the last step of a pursuit, and most work is won or lost before it is written. A conversion brain runs the decisions upstream of the document: which opportunities to pursue, on what strategy, saying what, to whom. Full comparison
Not a BD hire. A capable business development manager is valuable and expensive, and their judgement leaves when they do. A conversion brain captures judgement into an asset the firm keeps. Full comparison
Not pipeline automation. The systems built for software companies generate and enrich leads at volume. Engineering firms rarely have a volume problem; they have a conversion problem. Full comparison
The conversion brain is built on the method in Design Before Scale by Aidan Montague: fix the conversion design first, then scale it with AI. Montague is an engineer by background, mining and oil and gas, who later ran sales and marketing for major technology companies. The conversion brain treats conversion the way a design office treats a structure: specify it, engineer it, prove it under load, then let it carry weight.
The first install is deliberately modest. Thirty days: the current conversion bottleneck fixed before any AI is introduced, the Red Lines written, the first Winning Rules encoded, one workflow live with a human reviewing every output. Autonomy comes later, and only where it is earned.
Who is it for? Established firms selling complex services to technical buyers: engineering consultancies serving mining, energy, resources and infrastructure, and businesses like them. Start with the diagnosis, or read about the method.
Every install is built around a specific client-your business. Working out the fit and the shape will require an initial conversation (without obligation) and one or more follow-up scoping meetings.
Tell me a bit about your business using the form, and I’ll come back to you personally. We’ll look at where your conversion is leaking, where your hardest-won knowledge is locked up, and what the First Win would be for your team. No charge, no pressure. If it’s a fit, I’ll show you how the install starts. If it’s not, you’ll still walk away knowing where your real constraint is.
