Use case · Media plan
Spend your media budget where the AIs actually read
Every candidate for your budget arrives with an audience story. AI answers have their own reading list, and it can be measured: Epovest turns your prompts into the list of places the engines cite, with the weight of each. The split stops being a matter of conviction: the heaviest place takes the first allocation, and every placement gets a date, facing the curve it was bought to move.
Views from the app. Your figures replace the sample at the first survey.
Influencers, blogs, comparison sites: which ones feed the answers?
A media plan is a bet on where recommendations are made. When the recommendation comes from an engine, the bet can be replaced by a reading: every answer cites its sources, and keeping them turns your market into a ranked list of places, each with the share of your prompts it covers and its AI Authority per engine. Where and how much becomes an order: the weight of each place says where the budget starts, and the curve says, placement by placement, what moved.
What you measure
The places, ranked by weight
Every page the engines cite on your prompts, grouped by place: the share of your market each one covers, and its AI Authority, engine by engine. The comparison site, the blog and the creator channel read on the same scale.
The spending order
Sorted by weight on the engine you want to move, the list becomes a spending order: the largest allocation goes to the heaviest place, and a candidate the answers do not cite waits, instead of taking budget on faith.
Each placement, a dated proof
The day a placement goes live, the page is recorded as a corroboration, dated, with its channel. Weekly monitoring holds what you paid for: if your line leaves the comparison or the passage changes, you know the same week.
Renewal read on the curve
Each placement is an annotation on your citation series: what moved after it is your proof, and what did not move is your renewal conversation. The next allocation follows the re-read map, not the memory of the last one.
The recipe
How to build an AI media plan with Epovest
The full working guide for this use: your prompts turned into the ranked list of places that feed the answers, the budget cut into allocations that follow the measured weight, each placement bought, dated and monitored, and the renewal decided on the curve.
The flip: read before you sign
Every candidate for your budget claims a reach. The answers of the engines carry their own list of places, and it can be measured.
Start from the questions your buyers ask the AIs, not from the rate cards you received: a tracker on those prompts keeps every answer with every page it cites. The list of places that comes back rarely matches the one you were pitched, and that difference is the point: the plan is built on the first, and the second is read against it.
The map: every place with its coverage and its weight
The tracker keeps every page the answers cite on your prompts; a Competitor Scan widens the map with the names of your market.
Grouped by place, the citations become a map: how much of your market each place covers, and its AI Authority on each engine. Your candidates read on it directly: the comparison site cited across the market, the blog cited on two names, the newsletter, the channel. And the map is bigger than your shortlist: the places nobody pitched you are on it too, with the same numbers.
The split: the budget follows the weight
Sort the places by AI Authority on the engine you want to move: that order is the spending order, heaviest first.
Size the allocations on the two numbers of each place: the share of your market it covers, and its weight on the engine that matters to you. The amounts stay your decision; the order stops being one. A candidate the answers do not cite this quarter is not a refusal, it is a line that waits for the next reading. Each allocation you commit opens its quest, named after its place: the negotiation, the listing, the brief, worked as a file.
Each placement bought becomes a dated proof
The day the page goes live, record it as a corroboration, with its date and its channel.
A corroboration is the receipt the plan keeps: the page your budget bought, dated. Arm weekly monitoring on every paid page: a line that leaves the comparison, a passage that changes, and you know the same week, with a renewal conversation that carries proof instead of impressions. And every place a check surfaces enters your Atlas with its AI Authority: the map your budget works on is yours, and it keeps its history.
Renewal is decided on the curve
Every placement is dated in the Logbook; your citation rate and share of voice read before and after, engine by engine.
Keep the tracker running through the quarter: each placement becomes an annotation on the series the day you date it. At renewal, the questions have numbers: what moved after the comparison listing, what moved after the collaboration, which engine followed. The next allocation is split on the re-read map: heaviest place first, this quarter's numbers, and the candidates that earned their line since.
Worked example
One quarter, three candidates, one budget
A software vendor holds a quarterly placement budget and three pitches: a comparison site, a specialist blog, a creator channel. A tracker on eight buyer prompts, on ChatGPT and Gemini, plus a scan on four names of the market. The map returns: the comparison site covers most of the market with the heaviest AI Authority on ChatGPT, the blog carries two rival names, the channel is not on this quarter's map. Two allocations committed in that order, the third line held for the next reading. Both placements are live three weeks later, recorded as corroborations, monitoring armed, dated in the Logbook. The next checks read the share of voice against the two dates: the renewal is decided on the curve, and the held line is re-read on the widened map.
With your AI assistant
Your AI assistant runs this recipe with you
Connect Epovest to the assistant you already use, one-click connector or API key. It creates trackers, reads the results and keeps the registers with you; a payment only ever happens with your approval.
The instruction to give it
Connect to Epovest and read tracker [name]. Group the sources cited over the last 30 days into places, ranked by AI Authority on [engine], mark for each whether my name is there, and give me a split of [amount] across the five heaviest, heaviest first, with one quest per placement.
$0.10
per prompt, per engine, per check
Pricing
Whatever you configure, the unit does not change. Your total shows live in the configurator before anything runs; you start whenever you want, from a $10 top-up. No subscription: only the checks that run are debited.
Your budget changes, the map keeps its dates
Add an engine, reword a prompt, put a new candidate on the map when the quarter changes: every check keeps its date, so last quarter's map stays readable next to this one's. The monthly rescan widens the list of places rather than refreshing it: what one pass found, the next one keeps.
The other use cases
Make AIs recommend you
Create a tracker in minutes: your prompts, your engines, your cadence. The first check runs the moment you start.
No subscription. Credits from $10 that do not expire. Pause anytime: the history stays yours.