When you’re launching with a lean catalog, the fastest, least wasteful path is a two-stage system: Performance Max to discover winners, then Search to control cost & scale. It’s the same principle behind my ROI-positive builds and case studies.
Table of Contents
- Why PMax First, Search Second
- Quick Pre-Launch Checklist (feed, conversion tracking, policy)
- Stage 1: Performance Max for Discovery
- Stage 2: Search Campaigns for Winners
- Bidding, Budgets & Negatives
- Metrics to Watch Weekly
- When to Add Shopping & Remarketing
- Common Mistakes & Fixes
- FAQs
1) Why PMax First, Search Second
- Even exposure: PMax gives each product a fair audition so your early data isn’t biased by guesswork.
- Faster signal learning: With a small catalog, you’ll collect conversion data fast enough to identify winners.
- Then control: Move winners to Search where you decide bids, queries (exact match), copy, and extensions.
2) Quick Pre-Launch Checklist
- Product feed: Titles = Brand + Model + Key Attribute + Use Case; include GTINs, prices, availability.
- Tracking: Primary conversion = purchase. Pass value, currency, and ideally new/returning customer flags.
- Policies: Avoid disapprovals/suspensions by aligning site, payments, and claims with Google Ads policies. If you do get suspended, follow my step-by-step restoration guide.
3) Stage 1: Performance Max for Discovery
- 1 campaign, 1 asset group (per category if >10 SKUs).
- Budget: Start small but steady (e.g., 1–2× target CPA per day).
- Signals: Feed PMax your top search terms and website visitors; upload a seed audience if available.
- Run 10–14 days (no big changes). Tag every order back to SKU.
Graduation rule: Promote any SKU once it achieves ≥2–3 purchases at or below your target CPA/ROAS.
4) Stage 2: Search Campaigns for Winners
- Structure: 1 winner = 1 ad group; tight themes.
- Match types: Start with exact match for control; add phrase later if you have budget headroom.
- Bidding: Manual CPC (or eCPC) at launch for cost control; move to tCPA/tROAS after 15–30 conversions. This mirrors the “manual first, then automate” philosophy discussed in your talk.
- Ad copy: Emphasize key attribute + use case; mirror product title; include urgency (stock, delivery).
- Assets: Sitelinks (size guides, shipping/returns), price extensions, structured snippets.
- Negatives: Add non-buyers, cheap-intent queries, and competitor names you won’t bid on.
5) Bidding, Budgets & Negatives
- Start tight: Exact match + manual bids contain CPC while you test. (From the session: “manual bids & exact match” as you graduate winners.)
- Scale: Raise budgets on ad groups that hit CPA/ROAS goals for 3–5 days in a row.
- Negative hygiene: Review Search Terms 2–3×/week the first month.
6) Metrics to Watch Weekly
- Per-SKU: CTR, CPC, Conv. Rate, CPA, ROAS, Impression Share Lost (budget/rank).
- PMax: Conversion value by item ID; use that to pick promotions.
- Search: Query-level profitability and Quality Score components.
7) When to Add Shopping & Remarketing
- Standard Shopping: Use if you want additional query control alongside PMax, or to isolate new SKUs. (Optional path mentioned in your talk.)
- Remarketing: Build RSAs and Display remarketing lists once you hit 1k+ visitors/30 days.
8) Common Mistakes & Fixes
- Too many campaigns for too few SKUs → consolidate.
- Switching bids daily → let data mature (3–7 days).
- No negatives → CPC bloat; fix with weekly curation.
9) FAQs
- Why not start with only Search? You’ll bias spend to what you think will win, not what customers prove. PMax speeds up SKU discovery.
- Will this scale? Yes—see my case study results and approach.
- Where can I reach you? Contact page on my site.