Programmatic Display Guide for Retail & E-Commerce

If you sell on Amazon, Amazon DSP is essential for accessing Amazon's purchase intent data and extending reach off-Amazon. The Trade Desk is the strongest independent DSP for open web and PMP buying. DV360 offers deep integration with Google Shopping data. For mid-market brands, StackAdapt and Basis Technologies offer simpler UI and lower minimums.

Audit your programmatic supply chain to minimize intermediary fees. A typical open auction impression passes through 2–3 SSPs before reaching the DSP, each taking a margin. Work with your DSP to identify the shortest path to premium publisher inventory and set up preferred supply-path rules.

Allocate roughly 30% of display budget to open auction (scale and retargeting), 40% to curated PMP deals (quality inventory at reasonable CPMs), and 30% to programmatic guaranteed (premium placements with guaranteed delivery). Adjust based on brand safety requirements — direct-to-consumer brands with sensitive creative may skew 70%+ to PMP/PG.

Your most valuable targeting asset is your own customer data. Upload CRM and purchase history data through your DSP's identity resolution partner (LiveRamp, UID2) to build seed audiences. Create segments by recency (30/60/90-day purchasers), frequency (repeat vs. one-time), and AOV tiers.

As third-party cookies decline, contextual targeting is experiencing a renaissance. Use AI-powered contextual engines (Oracle Contextual, IAS Context Control, Peer39) to place ads alongside relevant editorial content — fashion editorial for apparel, product review sites for electronics, recipe content for kitchen/home goods.

Set up product-level retargeting using your product feed. Users who viewed specific products should see those exact items in display ads within 1–7 days. Configure the retargeting window by product category: impulse categories (beauty, accessories) at 3–7 days, considered purchases (electronics, furniture) at 14–30 days.

Suppress users who purchased in the last 7–30 days from prospecting and retargeting campaigns. Exception: for high-repurchase categories (consumables, beauty refills), shift recent buyers into a replenishment retargeting segment with messaging about their likely next purchase rather than suppressing entirely.

For dynamic product ads, feed quality is the single largest performance lever. Ensure product titles are descriptive (include brand, product type, key attribute), images are high-resolution on white backgrounds, prices are accurate and up-to-date, and out-of-stock items are automatically excluded.

Use static creative for upper-funnel brand awareness (lifestyle imagery, brand story) and dynamic creative for mid- and lower-funnel (specific product recommendations, pricing, urgency messaging). Static ads build brand; dynamic ads drive conversion.

High-impact formats (skins, adhesion units, interstitials) command $4–10+ CPM but deliver 5–10x higher attention than standard banners. Use these for product launches, seasonal tent-pole campaigns (Black Friday, Prime Day), and brand storytelling moments.

Display CTR averages 0.05–0.18% — optimizing toward CTR drives spend to bot-prone, low-quality placements. Instead, measure view-through conversions (with a conservative 1–7 day window), attention metrics (Active Attention seconds), and incrementality via holdout studies.

Reserve 10–15% of your display audience as a holdout group receiving no display ads. Compare conversion rates between exposed and holdout groups to measure true incrementality. Retail retargeting often shows true incrementality of only 20–40% of reported ROAS — this is critical intelligence for budget allocation.

Integrate return data into your attribution model. Products with high return rates (fashion: 20–35%) inflate gross ROAS while destroying net economics. Calculate net ROAS by channel and product category to avoid over-investing in return-heavy segments.