Choose an offer you can fulfil and a visual payoff you can prove. Pinterest’s wider fall report compares global text-based searches from July 2025 to July 2026, but publishes no absolute counts. These three examples are unranked test prompts, not a demand forecast.
Three selected report examples
| Search growth | Product fit | Visual proof |
|---|---|---|
| “dark red cat eye nails design” (+735%) | Stocked red magnetic polish or a bookable nail service. | Reflective pigment in close-up. |
| “mocha balayage hair” (+949%) | Warm brunette colour or bookable balayage service. | Warm ribbons against a darker base. |
| “marshmallow perfume” (+248%) | Sweet-profile perfume in stock. | Bottle in a warm, creamy setting. |
Worked hypothetical: compare motion
Assume a hypothetical nail brand has dark-red cat-eye polish in stock for the chosen market. Produce two six-second standard-width video ads in 9:16 vertical following Pinterest’s standard-width video specs, from one locked macro shot of the same hand and bottle. Keep copy, “Red that catches the light,” and CTA, “Shop the shade,” identical; match crop, hand, shade, background, audio, offer, landing page and campaign settings. The control holds the base frame for all six seconds. The variant adds one light sweep across the polish from seconds 2–4; keep every other frame and overlay identical. Both assets are videos of the same duration, so this isolates motion rather than comparing still image with video. This is a hypothetical brief; neither asset nor campaign exists.
Expected deliverables: two matched exports named control-static-hold and variant-light-sweep, plus a one-page setup and decision sheet. The copy and CTA are illustrative; the advertiser supplies the real SKU, market and offer.
Hold market, audience, campaign structure, objective, placements, flight window, offer, landing page and total budget constant; split 50/50 if supported. Preselect one metric (traffic: outbound CTR/clicks; sales: checkout CPA/ROAS) and a business guardrail. Use Ads Manager’s estimated power and regional minimum-budget guidance; Pinterest recommends 14 days to two months in most cases. Multi-ad results aggregate at test-version level. See Pinterest’s A/B test guidance.
Advance the variant only with adequate test power, a statistically supported improvement in the preselected metric and a guardrail pass. Weak or indistinct results are inconclusive; reject a clearly worse variant. Higher outbound CTR alone is not a sales win if checkout CPA exceeds the pre-set ceiling or ROAS misses its floor. No result is claimed.
Keep the ads easy to read on a phone, product-forward, clearly branded and action-oriented, in line with Pinterest’s creative guidance.
Align stock and offer timing with the broader holiday-marketing plan.
Check Visual Search Ads placement guidance separately before building a placement-specific version.
Beauty Guides have a separate rollout
On 6 October 2026, Pinterest’s Beauty Guides announcement says the feature is available to iOS users in the U.S. and plans to expand to more people and regions and additional product categories over time. It gives no date for broader access. This describes a feature rollout, not the report’s search geography.
