How we decide what work to chase
Most small studios pick their next market the same way: someone has a good feeling on a Tuesday, and six weeks later there is a landing page for it. We wanted something better than a good feeling, so we built the thing we would have built for a client. A register of where we might grow, what we might sell, and exactly who we would call.
The register has three layers, and the discipline is in keeping them separate. A strategy is a route to a set of buyers. An opportunity is one concrete play inside that route. A prospect is a named company with evidence attached. Most plans blur all three, which is how you end up with a market you like the look of and no idea who takes your call.
Every opportunity carries six scores from one to five: how binding the deadline is, how underserved the buyers are, how much of the delivery AI can carry, the size of the market, how well the engagement fits two senior people on a fixed price, and how much recurring work follows. Those six are weighted into a single number out of a hundred, and the number is calculated by the database from the six inputs rather than typed in by hand. That sounds like a technical detail. It is actually the whole point: a score you can type is a score you can talk yourself into.
The uncomfortable part is that the model is built to be unkind to our own enthusiasm. Deadline pressure carries the heaviest weight, so anything without a real clock starts at a disadvantage no matter how much we like it. One of our own offers, the one paying the bills today, scores mid-pack. We left it that way. A register that flatters you is just a mood board with arithmetic.
The six rules we kept coming back to
Distribution beats demand
Knowing what to sell is easy. Knowing who will take your call is not. The strategies that moved for us were the ones built on a real relationship, not the ones built on a big market.
Find the clock
The best work is work someone is forced to buy by a date they did not choose. A regulation, a vendor end-of-life, an insurance renewal. Nobody has to be persuaded the problem is real.
Score against yourself
A three is average, and most things are twos and threes. If every idea scores well, the model is broken. We keep the unflattering numbers because the register is sorted by them.
Name the company or say you cannot
A prospect is a real business with a reason attached, not a category. An empty contact field is a correct answer. A plausible invented one costs a week to discover it was wrong.
Write down what would kill it
Every entry records its risks and what remains unverified. The assumptions we cannot check yet are the most valuable thing in the file, because they tell us what to test first.
Drafts are drafts
We use AI to research and draft candidates, and nothing it produces is saved until a person reads it and accepts it. The rejected drafts are kept too, along with the prompt that caused them.
The method earned its keep almost immediately, and not in the way we expected. The single largest market we looked at, worth tens of millions on paper, came out with one of the lowest scores in the register. Seven well funded competitors already run sales teams against that exact list, and we have no way in. So it sits there, written down, scored honestly, and parked. Later, from a completely different direction, we found a route into the same buyers through someone who refers them work every week. Same market, same offer, different door, and now it is worth doing.
That is the argument for writing this down rather than carrying it in your head. Ideas do not get discarded, they get scored and parked, and when the missing piece turns up you already know exactly what it unlocks.
We build this kind of thing for clients too. Not a dashboard, and not a strategy deck that ages badly. A working system that makes the next decision easier than the last one.
Want to know where your next quarter should go?
We settle the thinking first, then build the system that runs on it. Start with a free chat.