AI you built in-house will not sell: pick one budget line they fund
Before getting the market to learn a new category name, pick one budget line the buyer funds this year.
Key points
- No budget line on the buyer side
- Ask who applies what last year funded
- The line sets the rival and the anchor
- Spread a new name only if no line fits
Overview diagram in preparation
Contents
It stalls for want of a budget line, not for want of explanation
A&A perspective
A budget line is a category a buying company has already assigned spending to; in accounting terms, an expense account. Every company has lines carried over from last year: IT maintenance, outsourced sales work, training.
A&A perspective
Sometimes the AI setup you built for your own company turns out to be sellable to others. You show it, the response is good, and the buyer says it is needed. Then they take it back inside, and the thread goes quiet.
A&A perspective
The silence is not disagreement. Applying internally for the use of budget is what we will call applying here. There is no line on the buyer side that the application can use. To get sign-off, the person there picks one existing line and writes the amount into it. With no line to pick, the work stops.
A&A perspective
What most people add here is explanation. The deck gets thicker, the words get swapped, more examples go in. The assumption is that explanation is what is missing.
A&A perspective
Automate & Augment (A&A), the company writing this article, takes on rebuilding how work gets done on the assumption that AI is in the loop.
A&A perspective
A&A treats this as a sign-off problem, not a comprehension problem. More explanation does not add a line inside the buyer. A public record of the same stall exists overseas.
From the sources
Clay, which provides tooling that automates sales research, publishes a customer story about Lovable. Lovable is a service that builds web apps from natural-language prompts.
From the sources
The story names one of the problems to solve this way. The "vibe coding" category didn't exist before.
From the sources
The category of building apps by prompt did not exist until then, is how it is written. The same passage continues.
From the sources
"There was limited established market understanding of what an AI app builder could do or who it was for".
From the sources
What an AI app builder could do, and who it was for: the understanding the market had built about that was limited, is what it says.
A&A perspective
What is named as the problem there is not the explanation of what the product can do. It is understanding on the market side. When a small company sells, A&A reads that missing understanding as showing up as a missing line on the buyer side. The buyer follows the words; what is missing is only the line to write the amount into.
Decide the line by asking who applies what last year funded
A&A perspective
The line you ride is not decided by your proposal. It is decided by which slots exist inside the buyer.
A&A perspective
A&A uses one rule. Look at the slots funded this year inside the buyer, and pick the one slot where the result of this work lands first. Slots created from next year, and slots that might be created on request, do not count.
A&A perspective
Ask for the candidate lines before you send the proposal. Ask the person who applies, not the approver. That person is the one who remembers which line bought what last year.
From the sources
On fitting yourself to one company at a time, there is a passage in a piece an entrepreneur published in 2013, written for early founders and urging them to do unscalable things by hand.
From the sources
"The most common unscalable thing founders have to do at the start is to recruit users manually".
From the sources
The most common unscalable work a founder does first is recruiting users by hand, is what it means. On business-to-business dealings, the same piece goes further.
From the sources
"pick a single user and act as if they were consultants building something just for that one user".
From the sources
Pick one company and behave like consultants building for that company alone, is how it is written. The same passage continues: "The initial user serves as the form for your mold".
A&A perspective
The first user becomes the form your mold is taken from, is what it means. A&A reads that route as a way to find the line. Fitting the work to one company reveals which slot that company filed it under. The piece says nothing about budget lines or sign-off; this re-reading is A&A's own.
Hypothetical example
What follows is a hypothetical. Say you sell a setup that moves internal enquiry handling onto AI. The candidate lines are IT maintenance, outsourced sales work, and training.
Hypothetical example
If maintenance is the only slot funded this year at that buyer, that is the line. This is an invented example, not a record of an actual engagement.
| Candidate budget line | Who you are compared with | Price anchor in that slot |
|---|---|---|
| IT maintenance | the company holding maintenance now | the existing maintenance fee |
| Outsourced sales work | an agency | the agency retainer |
| Training | a training firm | the per-session rate |
Choosing the line also chooses the rival and the price anchor
A&A perspective
Picking a slot costs you something too, because the slot already has an occupant.
A&A perspective
Ride IT maintenance and you are compared with the maintenance fee of the company holding that contract now. Ride outsourced sales work and you sit beside an agency retainer. Ride training and the anchor becomes a training firm's per-session rate.
A&A perspective
So choosing a budget line is choosing who you are compared with and which price anchor applies. Neither source says this; it is A&A's reading.
A&A perspective
There is a standing objection: invent a new name so you are not beaten down on price. It needs an answer here.
A&A perspective
A&A's answer is this. On the naming road, being compared still happens. Only the timing changes. The comparison slides later, and while it slides you pay for teaching the market yourself.
A&A perspective
On the budget-line road, the rival and the price anchor are visible first. What is visible can be built into the proposal.
A&A perspective
How you build it in changes with the slot. On maintenance, raise a line item for what the current maintenance scope leaves out. On outsourced sales work, raise a line item for the volume of work you replace.
A&A perspective
Either way, you add your line on top of an anchor the other side already holds. The material so far fits on one sheet.
A&A perspective
List three candidate slots, and write three things on each row. Whether it is funded this year. Who you are compared with. What the price anchor is in that slot.
A&A perspective
A&A has not measured whether the filing line changes win rate, price or days to sign-off. It offers no ranking of which lines are better.
A&A perspective
This is as far as it goes. Choose the slot, and the rival and the price anchor are decided with it.
Take the naming road only when no line fits at all
A&A perspective
When not one slot can be found, the other road remains: invent a new name and get the market side to learn the category.
From the sources
The Lovable story took that road. What it teamed up with were creators who demonstrate the product in action. The same page gives this reason.
From the sources
The route where the money spent works hardest, for getting the product understood and for making people want it, is how it is written. Teaching the market is written as the purpose itself.
From the sources
The same page also writes the load of that route: "the universe of relevant creators is large, the quality varies enormously".
From the sources
Relevant creators are many and their quality varies widely, is what it means. The teaching route is not something you get for free.
From the sources
The scale of the side that took that road is on the same page: "Over 40 million projects have been built on the platform since launch".
From the sources
More than 40 million projects have been built on that service since launch, it states.
A&A perspective
A&A takes one view from this. The road of teaching the market is open to businesses that can pay for the teaching themselves. The source only writes the scale and the load of the route; it does not say which scale cannot take it. That is A&A's reading.
A&A perspective
Scale is not the only reason it does not transfer. The creator route works because a viewer can try the product on the spot. A way of selling where a person is attached to each delivery has no such route. That too is A&A's judgement.
A&A perspective
The same page also carries a result figure, "+50% more qualified meetings per rep": 50% more qualified meetings per sales rep. No denominator, period or definition is given, so A&A does not use it as a forecast for its own work.
A&A perspective
So for a small company, the naming road is not the default. Write out three candidate lines, and keep it as the exception that remains only once none of them fits.
Summary
A&A perspective
It stalls because there is no budget line on the buyer side, not because the explanation is thin.
A&A perspective
Decide the line by asking the person who applies what last year funded.
A&A perspective
Once the line is decided, the rival and the price anchor come into view.
A&A perspective
The road of spreading a new category name remains only once you have confirmed that no line fits.
A&A perspective
A proposal sent after one line is decided can move straight into the buyer's internal application. That is the one thing decided before the explanation gets thicker.
A deal stalls inside the buyer because there is no budget line the internal application can use, not because the explanation is thin. Decide the line by asking the person who applies what last year funded. Once the line is decided, the rival and the price anchor are decided with it, and both can be built into the proposal. The naming road remains only once you have confirmed that no line fits.
FAQ
What if, after fixing one line, a different line turns out to be easier to pass?
Move it. The line is an attribute of the proposal, not of what you deliver. Moving it swaps the rival and the price anchor too, so rewrite the line items to match. The point of deciding is to go out with one.
What if the buyer is a one-person company with nobody who applies?
The person you ask becomes the owner; what you do is unchanged. Ask which lines are funded this year and what each line paid for last year. Even a one-person company has last year's spending categories to hand.
Is the naming road completely closed to a small company?
It remains once you have confirmed that no line fits. What this article advises against is taking it as the default before checking. The source shows the scale of the side that took it and the load of the creator route; it does not say which scale cannot take it.
Sources & editorial note
Primary pages read for this article. Publication dates below belong to the sources; access dates record our research.
- How Lovable uses Clay to scale one of the fastest-growing startups in history
Clay · undated customer story page; no publication or update date shown
Accessed 2026-10-25 - Do Things that Don't Scale
Paul Graham · July 2013
Accessed 2026-10-25