Prospecting vs retargeting
Prospecting is cold outreach: new audiences, lookalikes, broad targeting, people who don't know your brand yet. Retargeting is warm follow-up: site visitors, cart abandoners, past customers, email/SMS subscribers, people who've already shown intent. They don't compete for the same job.
Prospecting builds the top of the funnel (expensive, lower ROAS, because you're paying to introduce yourself). Retargeting harvests the bottom (cheap, high ROAS, because the buyer already decided you're worth considering). But it can only re-reach people prospecting already found, so alone it plateaus.
AI can classify your campaigns and compute the true split. The spend and ROAS numbers must come from your own Ads Manager, never a guess.
Fastest path: one prompt, end to end
🤖 AI prompt — paste into ChatGPT / Claude
You are my funnel analyst. Here is my campaign list with 30-day spend and ROAS, pasted from Meta Ads Manager: [paste: campaign name | audience type | spend | revenue | ROAS].
1. Classify EACH campaign as prospecting (cold: new/lookalike/broad) or retargeting (warm: visitors, cart, customers, lists).
2. Sum spend in each bucket and compute my true prospecting / retargeting split (% of spend).
3. Show blended ROAS vs the two bucket ROAS separately, and flag if a high blended number is being propped up by a shrinking retargeting pool.
Output ONE markdown table: Campaign | Bucket | Spend | ROAS, then a summary line: prospecting % / retargeting % / prospecting ROAS / retargeting ROAS.
Use ONLY the numbers I pasted. Do not guess, invent, or fill in a missing campaign; if a row lacks spend or ROAS, leave it blank and ask me. Never treat blended ROAS as either bucket's ROAS.
Or do it in 5 steps
- Build two buckets in Ads Manager. One set of prospecting campaigns (broad, lookalike, interest) and one set of retargeting campaigns (site visitors, cart abandoners, customer/email lists). Name them so the split is visible at a glance.
- Start at ~70/30, prospecting-heavy. Roughly 70% prospecting / 30% retargeting early on. Under ~500 orders/month, lean closer to 85/15: you don't have enough warm traffic yet to justify more retargeting spend.
- Segment retargeting by intent. A homepage browser and a cart abandoner are different levels of intent, so give them different messages and frequency, not one blanket "everyone who visited" audience.
- Set frequency caps and exclusions. Cap retargeting around 4-6 impressions/week (fatigue sets in past 7-8) and prospecting around 3.5. Exclude recent purchasers and existing warm audiences from prospecting so you don't pay cold prices to reach people you already have.
- Judge each bucket on its own metric. Prospecting on cost-efficient reach and new-to-brand rate; retargeting on ROAS and conversion rate. Recheck the split every 4-6 weeks as the warm pool grows.
Done looks like: two clearly named buckets, an intended split, frequency caps set, and purchasers excluded from cold campaigns.
Worked example: why blended ROAS misleads (illustrative math)
| Bucket | Spend | Revenue | ROAS |
|---|
| Prospecting | $8,000 | $16,000 | 2.0x |
| Retargeting | $2,000 | $12,000 | 6.0x |
| Blended | $10,000 | $28,000 | 2.8x |
The blended 2.8x looks healthy, but cut prospecting to "chase" the 6x and the retargeting pool stops refilling. Within weeks the 6x bucket has fewer warm people to reach, its revenue falls, and blended ROAS drops below where it started.
The two buckets earn their keep differently; read them separately. Numbers here are illustrative, run yours through the prompt above.
Re-check the split every 4-6 weeks.