Backed by Y CombinatorTwo documents and three sheets that find campaigns bidding against each other, and prospecting ads still running at people who already bought.
Free download · No account needed
What you get
Pack of 5 files: 2 documents, 3 spreadsheets
Paid Social Audience and Exclusion TemplateExclusion Map
Checked against Audience Register every refresh cycle. A blank cell reads as unchecked, not as no.
| Prospecting Audience | Customers Excluded? | Open Opps Excluded? | Closed-Lost Excluded? |
|---|---|---|---|
Any audience missing the active-customer exclusion gets fixed the same day it is found.
Two campaigns targeting the same person are not free just because nothing is technically double-billed. Meta's own documentation says when two of an advertiser's own ad sets can reach the same person, Meta enters only the stronger performer and suppresses the other, specifically so an account does not bid against itself. That protects the budget from paying twice for one impression. It does not mean the suppressed ad set is testing what its own report appears to show, because a real share of its intended audience was never eligible to see it.
On this pack's own worked account, two Meta lookalike audiences built from the same customer seed list overlapped 64%: almost two-thirds of the smaller audience's reach sat inside the larger one, both eligible for the same auction. A third prospecting audience, built on interest targeting before any exclusion list existed, carried zero exclusions of any kind. Separately, the active-customer exclusion list ran 9 weeks past its weekly refresh cadence, long enough for 306 newly closed customers to stay eligible for cold prospecting ads, an estimated $64.78 in spend shown to people who had already bought.
Overlap Matrix checks prospecting audiences against each other before a new one launches, not after a disappointing first week. Exclusion Map states, for every prospecting audience, which of three exclusions actually applies: active customers, open opportunities, or closed-lost. A gap reads as a decision or an oversight, never a mystery. The audiences behind both sheets often come from the same creative performance export already pulled for testing, and the spend they cost settles into the cross-platform paid report once the exclusions are fixed.
Audience RegisterSample rows for a fictional company's Meta and LinkedIn accounts.
| Audience | Type | Size | Cadence | Days Since Refresh | Status |
|---|---|---|---|---|---|
| Lookalike 1% (customer seed) | Prospecting | 42,000 | Monthly | 18 | On schedule |
| Interest: software category | Prospecting | 96,000 | Quarterly | 41 | On schedule |
| Active customers | Exclusion | 2,140 | Weekly | 63 | STALE (9 wks over) |
| Open opps, Proposal+ | Exclusion | 190 | Weekly | 9 | STALE (2 days over) |
A stale exclusion audience is still "applied" to campaigns, but applying an outdated list, missing everyone added since.
Overlap Matrix| Audience A | Audience B | % of A in B | % of B in A | Same Auction? | Flag |
|---|---|---|---|---|---|
| Lookalike 1% | Lookalike 3% | 64% | 21% | Yes | High - exclude A from B |
| Lookalike 3% | Interest: software | 29% | 39% | Yes | High - consolidate |
| LinkedIn job titles | Lookalike 1% | n/a | n/a | No | Different platform |
64% of the 1% lookalike's reach already sits inside the 3% lookalike, both eligible for the same Meta auction.
Exclusion Map| Prospecting Audience | Customers? | Open Opps? | Closed-Lost? | Gap Flagged |
|---|---|---|---|---|
| Lookalike 1% | Yes | No | No | Missing 2 of 3 |
| Interest: software | No | No | No | CRITICAL - built pre-exclusion, zero applied |
| LinkedIn job titles | Yes | Yes | No | Missing 1 of 3 |
The active-customer exclusion is never optional on a live prospecting audience. Every other gap here gets a stated decision, not a silent default.
How prospecting and exclusion audiences fit together, and what Meta's own auction actually does with two of your own overlapping ad sets.
The refresh cadence for each exclusion type, and the one exclusion, active customers, that is never optional on a live prospecting audience.
Every audience in use, its platform, source, size, and refresh cadence, with anything overdue flagged rather than left to be noticed.
Which same-platform prospecting audiences are eligible to reach the same people, and by how much, before a new audience launches blind.
For every prospecting audience, which of three exclusions, active customers, open opportunities, closed-lost, actually applies.
Open the pack and send your current audiences and CRM exports, or download the Word documents and CSV sheets and build the register yourself.
List every prospecting, retargeting, and exclusion audience across platforms, with source, size, and how often each should refresh.
Compare same-platform prospecting audiences against each other, and flag anything over 25% for consolidation or an ad set-level exclusion.
Map every prospecting audience against the three exclusion types and fix the active-customer gap the same day it is found.
Yes, and the download is not cut down. The zip holds all five files in Word and CSV, no signup and no card. Edit with AI is the other branch: send your current audiences and CRM exports, and River builds your own Audience Register and Exclusion Map. More packs in the template library.
Word documents (.docx) for Audience Strategy and Exclusion Policy, and CSV (.csv) for Audience Register, Overlap Matrix and Exclusion Map, all in one zip. They open natively in Word, Pages, Google Docs, Excel, Numbers and Sheets, so nothing needs converting.
No, and that is the easy misreading of it. Meta's own system suppresses the weaker-performing ad set rather than double-charging for one impression. The real cost is that the suppressed ad set is not testing what its own report seems to show, since a real share of its intended reach was never eligible to serve at all.
The creative performance analysis tool works out which creative, and which attributes of it, are winning inside a given audience. This pack works one level up: whether the audiences themselves are set up sensibly, without overlap or a missing exclusion, before creative performance is judged inside them.
Active customers, yes, always, per Exclusion Policy's floor. Open opportunities and closed-lost are a real judgment call some campaigns make deliberately, such as a planned re-engagement of an old closed-lost segment, but that has to be a stated decision on Exclusion Map, not an audience nobody checked.
It creates a free account, installs these five files as a private workspace, and opens by asking which platforms you run paid social on and what audiences are currently live. Then it builds your own Audience Register and Exclusion Map from whatever exports or screenshots you send.
Written by
Co-founder & CEO, River
Chandler has spent the last six years working in AI, building tools for millions of users and Fortune 100 companies. In 2025, he co-founded River to build a new kind of operating system for AI agents.
Send your current audiences and CRM exports. River checks for overlap between your own campaigns and gaps in who gets excluded.
Edit with AI