Retail Media Search & Keyword Strategy

Example output

Retail media search & keyword strategy — Quill Valley Granola

Keyword segmentation
- Branded (defend): "Quill Valley," "Quill Valley granola," and close misspellings. Small volume, cheapest clicks, highest conversion — and currently being encroached on by a competitor bidding your name.
- Category: "granola," "healthy granola," "granola clusters," "breakfast granola" — your biggest volume opportunity and the segment worth the most testing budget.
- Competitor: named competitor granola terms, used cautiously — a defensive/opportunistic play, not a volume driver; expect a lower conversion rate here than on your own branded terms.
- Long-tail: "gluten-free granola clusters," "low sugar granola," or whatever specific attributes your hero SKU actually carries — lower volume per term but often the cheapest, highest-intent clicks once identified.

Match-type & bidding approach
- Branded: exact match, aggressive bid — this is the cheapest defense against the competitor bidding on your name, and losing this auction is the highest-priority problem to fix first.
- Category: a mix of phrase and exact on proven terms; start a small broad or auto component specifically to discover new converting long-tail terms, then graduate winners out of broad and into exact.
- Competitor: exact match only, modest bid, monitored closely for conversion — cut fast if it's not converting within a defined test window.
- Long-tail: exact match once identified through the harvesting process below; these terms usually support a lower bid for a similar or better conversion rate.

Negative-keyword approach
Apply negatives aggressively on the category broad/auto campaign specifically — any search term that has spent budget across a meaningful sample size without converting should be added as a negative rather than left to keep bleeding spend. Review this weekly while the discovery campaign is active, since this is where a "modest monthly test budget" gets wasted fastest if left unmanaged.

Search-term harvesting workflow
1. Let the category broad/auto campaign run and report actual search terms.
2. Weekly, pull the search-term report and identify terms converting above your target rate.
3. Add winning terms to a dedicated exact-match campaign at a bid you control.
4. Add non-converting terms as negatives in the broad campaign so it stops re-learning the same lesson.

Budget split (directional)
Given the stated brand-term defense problem, weight budget toward branded defense first despite its low overall spend need — it's inexpensive to defend and expensive to lose. Roughly: majority to category (your growth engine), a meaningful minority to branded defense, a small test allocation to competitor terms, with long-tail folded into category once harvested. [Insert your actual monthly budget to set real dollar caps per segment.]

Measurement
- Primary: ROAS against a breakeven tied to your actual margin — [insert margin] to calculate this precisely.
- Brand-term impression share and auction win rate — the direct measure of whether the competitor-bidding problem is resolving.
- New-to-brand order share on category and long-tail terms, since branded-term conversions are largely existing buyers, not growth.
- Search-term report health: declining wasted spend on non-converting broad terms over time is a sign the harvesting workflow is working.

Want me to build the actual starting keyword list and bid recommendations for the category segment?