How to Turn a Category Map Into Prospects

Turning a category map into prospects means reading a market ranking as a target list. The brands at the bottom of a ranking aren't failing at marketing. Most of them have no idea the ranking exists, which is the entire pitch. Read the ranking, pick the brands buyers know that AI rarely names, add them to a prospect list, and send each one a report about their own market before you ask for the call.

It's four steps, and three of them are judgment rather than clicks.

New-business work in this channel has an awkward first move. You want to open with a prospect's own data, and their own data is exactly what you don't have before they've signed anything. So most outreach falls back to a capabilities deck, which the prospect has already read twice this month from two other firms.

A tracked category fixes the sequencing. You measure the market first, and the market is public. By the time you call anyone, you can tell them who AI names when buyers ask about their category, where they sit in that list, and which publishers the answers are coming from. None of that required their permission, their analytics, or a single meeting.

This guide works with any AI visibility setup; the callouts show how each step runs in Spyglasses.

Step	What you're doing	What tells you it's working
1. Rank the brands	Reading the market ranking top to bottom	Five or more real brands, with the tail visible
2. Pick the gaps	Matching name recognition against share of answers	A shortlist of brands with something to show
3. Add the prospect	Moving a name into the prospect list	A property of its own, separate from your clients
4. Pitch the report	Leading with one finding, branded	A report in their inbox before the call

Rank the brands

Start with the map you already have. If the market isn't tracked yet, that's the prerequisite, and mapping a category in AI answers is the five-step version of it.

Read the ranking from the top down and note three things.

The share each brand holds. Share of answers is the percentage of completed answers a brand appears in, so a brand at 14% appeared in fourteen out of every hundred answers. The values don't add up to 100, because one answer usually names several brands and plenty of answers name nobody. Read them against each other, not against a whole.

Where the tail drops off. Some markets have a flat ranking, with eight brands inside a few points of each other, and some have a leader at 30% and nothing above 6% after that. A flat market is a harder sell and a steeper one is easier, because in a steep market the gap is obvious to the prospect on sight.

How long the list is. A market where AI names thirty brands is behaving differently from one where it names seven, and the seven-brand market is the one where being absent hurts most.

Pick the brands with a gap

This is the step that decides whether the list is any good, and it's a judgment call rather than a filter.

A gap is the distance between how well buyers know a brand and how often AI names it. A brand everybody in the market recognizes that AI mentions in 3% of answers has a gap. A brand nobody has heard of sitting at 3% has a small business, which is not the same thing and not your pitch.

So go through the ranking with two columns in your head.

Known by buyers	Named by AI	What it means	Worth a call
Well known	Rarely named	A real gap, and a visible one	Yes, this is the list
Well known	Often named	Already winning the channel	Rarely, unless they want to defend it
Barely known	Rarely named	A small brand, not a visibility problem	Usually not
Barely known	Often named	Somebody is already doing this work	No, they have a partner

That last row is worth pausing on. A brand punching far above its recognition is nearly always being helped by somebody, and walking in to tell them they have a problem they solved last quarter is the fastest way to end a conversation.

Two more filters before the shortlist is done. Drop anyone whose ranking is already climbing steadily in the 7-day changes, for the same reason. And drop anyone you can't actually serve, because a pitch list is only useful at the length you can work in a month.

Add one as a prospect

Now move the shortlist out of your head and into a list that does something.

Copy the brand name and the domain straight off the ranking. Add them to your prospect list rather than to your client roster, because a target brand you're researching and a client you're reporting on should never sit in the same view. Six weeks from now, when somebody else on the team opens the dashboard, that separation is the difference between a clean account list and a confusing one.

Keep the batch small. The temptation with a seventeen-brand ranking is to add all seventeen, and then you have seventeen reports nobody reads and a list that's stale by the time you get to the bottom of it. Add the ones you can call this month, work them, and come back.

One practical note on plans. Prospect properties come with an Agency plan and a trial doesn't include them, so this is the step where the workflow stops if the account isn't on one.

Pitch with a report they can read

Run a report on the prospect, read it yourself first, then send it.

Reading it first matters more than it sounds. The report will have a dozen things in it, and your job is to find the one that stings. Usually that's a named competitor. "When buyers ask who does this, AI names these four companies, and you're not one of them" is a sentence a prospect remembers through a whole meeting. A low number on its own isn't; it's a metric, and they've been shown metrics before.

Lead with one finding, not five. The rest of the report is the second meeting.

Send it ahead of the call rather than presenting it live. A prospect who's already opened a report showing their competitor winning their category shows up with a question instead of a blank stare, and that question is the one you wanted them to ask. Branded exports are what make this land, because a polished PDF with your name on it and their data inside reads as evidence you already understand their problem.

The category itself is the second half of the pitch, and it's the part other firms won't have. You're not just showing them their number; you're showing them the whole market, who's moving in it, and which publishers AI already reads for it. That last part goes straight into the first ninety days of work, whether it's earned media or content.

The map keeps working after the pitch

The useful thing about prospecting off a category is that the asset doesn't get used up. A capabilities deck is spent the moment you send it. A tracked market keeps running, and every week it tells you which brands moved, which new names appeared, and which publishers the answers are leaning on.

That means three things for a new-business pipeline. Your list refreshes itself, because a brand whose share slid four points this month is a warmer call than it was last month. New entrants announce themselves the first day AI names them. And when a prospect finally signs, you already have the market context to run their first quarter, which is a much better starting position than a kickoff call.

If you're building the wider prospecting motion, prospecting new clients with AI visibility reports covers running a batch of brand-level reports without a category behind them. Once a prospect becomes a client, baselining their AI visibility is the first week's work, and the category keeps running alongside their brand rather than being replaced by it. For what an agency plan covers, including prospect properties and white-labeled exports, see Spyglasses for agencies.

Part of the Spyglasses workflow series. This guide is one of eight workflows for marketing agencies, and it also appears in the guides for PR and communications teams. The rest are collected in the AI visibility guides for marketing agencies.