Audiences belong to the client, not to a campaign, so they sit on the client profile in a new Targeting tab, next to the Library. Built once, then picked in any brief, media plan or live campaign. Negative keyword lists live in the same tab (see the Search toolkit concept).
| Audience | Type | Meta | TikTok | Microsoft | Used in | |
|---|---|---|---|---|---|---|
| Current applicants Uploaded 2 Oct · 1,840 people | Customer list | Ready | Ready | Ready | Ready | 3 campaigns, excluded |
| Alumni 2018-2024 Uploaded today · 12,400 people | Customer list | Matching, up to 48h | Matching | Ready | Matching | Not used yet |
| Course page visitors, 30 days Website rule · pages containing /courses | Website visitors | ~6,200 | ~4,900 | Filling | Tag not found | 2 campaigns, targeted |
| Lookalike of current applicants Similar people in the UK · 3% | Lookalike | Signal only | Ready | Ready | Not offered | 1 campaign |
Each audience is created separately on every platform, so status is shown per platform. Hover a status for what it means. How ad platforms are structured
The user uploads a spreadsheet of existing customers. Emails and phone numbers are scrambled (hashed) before they leave the browser, and Snifferdog™ keeps only each platform's audience ID, never the list.
Use it to reach people who already know you, or to stop showing ads to people who've already applied.
Retargeting audiences are rules, not lists. Snifferdog™ offers simple presets and translates each one into every platform's own rule. They only work if the tracking tag is firing, so the tag check from the pre-launch checklist appears here too.
Show ads to people who've already been on calderbrook.ac.uk.
One “Lookalike” button, but it behaves differently per platform, so the screen says exactly what each one will do. Only audiences big enough to use can be a source.
Find new people who are similar to an audience you already have.
Audiences are picked, not built, inside a campaign: in the media plan before launch, and in live edit mode afterwards. Same panel in both places. Each audience can be targeted or excluded.
| Audience | Size on Meta | |
|---|---|---|
| Alumni 2018-2024 | Still matching | |
| Course page visitors, 30 days | ~4,900 | Added |
| Current applicants | 1,610 matched | Excluded |
| Audience type | Google Ads | Meta | TikTok | Microsoft Ads |
|---|---|---|---|---|
| Customer list | Customer Match. Eligibility rules mean some new or low-spend accounts can't use it. | Custom audience from a customer list. Terms must be accepted once per ad account, possibly inside Meta. | Custom audience (customer file). | Customer list. |
| Website visitors | Remarketing list from the Google tag. | Website custom audience from the Pixel. | Pixel audience. | UET remarketing list. |
| Lookalike | No lookalikes for Search (similar audiences retired in 2023). Lookalike segments in Demand Gen only. Lists act as signals in Performance Max. | Lookalike audience, 1 to 10% of a country. | Lookalike audience, Narrow, Balanced or Broad. | TO CONFIRM what's available. |
| How it's used | Search: bid adjustments (observation). Performance Max: signal. Demand Gen: targeting. | Targeting or exclusion. | Targeting or exclusion. | Bid adjustments or targeting. |
| Minimum sizes | TO CONFIRM Every platform has a minimum matched size before an audience will serve, and they differ. Read them from the APIs and show “Filling” or “Too small” rather than hard-coding. | |||
Customer lists are personal data even when hashed. Hashing protects the data in transit; it doesn't make it anonymous. The client is the controller and must have a lawful basis to use the list for advertising; Snifferdog™ is processing it on their behalf.
Written 9 Oct 2026 from the API difficulty sheet. Platform rules change, so re-check before the build starts.