Conversational commerce is not a distant hypothesis. It is already changing how some users discover and purchase products, and the gap between brands that are ready and brands that are not is starting to matter. Copilot Offer Highlights and Copilot Checkout represent Microsoft's current implementation of this shift — AI that does not just recommend products but can surface specific offers and facilitate the transaction.
The brands that benefit from these formats earliest are not necessarily the biggest. They are the ones with the most reliable product data, the clearest offers, and the cleanest merchant signals. This guide covers what those requirements actually look like in practice.
Why conversational commerce changes the pre-click experience
Traditional search advertising gives brands control over what shows up when a user searches a specific query. The ad copy, the extension, the landing page — these are all brand-controlled. Conversational commerce shifts control upstream. The AI reads the user's intent in natural language, synthesizes relevant options, and presents a curated recommendation. The brand's control is now primarily at the data layer: how complete, accurate, and structured is the product information the AI has to work with.
What Offer Highlights and Copilot Checkout actually do
Offer Highlights surface promotional information — specific discounts, bundle offers, or value signals — within Copilot's shopping recommendations. When a user asks for the best option in a category, Copilot can weigh offers from retailers whose promotional data is available and accurate.
Copilot Checkout goes further: the user can complete a purchase without leaving the Copilot experience. This compresses the purchase funnel significantly. The comparison, selection, and transaction all happen within the AI interface. Retailers who are not structured to participate in this flow are excluded from a meaningful portion of purchase decisions.
What retailers need in place before these formats matter
Product data quality
The foundation is a complete and accurate product feed. Every required attribute must be populated: GTIN, brand, condition, price, availability, high-resolution images, detailed descriptions. Missing attributes reduce confidence in AI recommendations. Mismatched attributes — particularly price or availability differences between the feed and the live site — can result in product exclusion from Copilot surfaces.
Offer clarity
Promotions must be structured clearly for machine consumption. Vague promotional messaging like "great deals this week" does not translate to AI offer highlighting. Specific, structured offers — percentage discounts, bundle pricing, free shipping thresholds — that are reflected accurately in promotional feed attributes are what the AI can surface.
Policy cleanliness
Returns policy, shipping terms, and customer service standards need to be clear, structured, and consistent across the Microsoft Merchant Center account and the live site. Copilot commerce features require that users can trust the post-click or in-experience transaction. Policy ambiguity or mismatches between what is listed and what is actually offered will prevent or limit participation.
Merchant trust signals
Microsoft evaluates merchant trustworthiness signals including account history, policy compliance record, and customer experience metrics. A Merchant Center account with a history of disapprovals, policy warnings, or customer experience issues will be deprioritized for premium AI commerce placements.
Where many brands will fail first
| Failure point | What it looks like | Fix |
|---|---|---|
| Stale price data | Feed price differs from live site after a sale ends | Automated feed refresh every 4–6 hours minimum |
| Missing GTINs | Products without manufacturer identifiers excluded from AI recommendations | Source GTINs from manufacturer or apply for custom labels |
| Vague promotions | Offer data present but too generic to surface | Structure promotions with specific values in feed promotion attributes |
| Returns policy gaps | Returns policy not specified in Merchant Center account settings | Add structured returns policy to account settings |
| Low-quality images | AI recommendations deprioritize products without high-res clean imagery | Minimum 1000x1000px, white or neutral background, no text overlays |
How this changes creative strategy
Traditional ad creative strategy focuses on writing compelling copy for a human to read. Conversational commerce creative strategy shifts emphasis to structured data that an AI can interpret and present accurately. This does not mean abandoning compelling copy — it means ensuring the product data layer is as strong as the creative layer.
Product descriptions need to be specific and factual enough to serve as source material for AI-generated recommendation copy. Titles need to include the specific attributes users search for: brand, model, key specification, size. Generic descriptions are not citable and not recommendable.
How this changes landing page strategy
For Copilot Checkout specifically, the transaction may bypass the landing page entirely. This means landing page optimization remains important for traditional search and Shopping traffic, but the conversion tracking and revenue attribution models need to account for transactions that occur outside the site. Ensure your analytics and conversion tracking are set up to capture order data from Copilot Checkout transactions in addition to on-site purchases.
Ecommerce example: apparel
An apparel retailer with seasonal promotions finds that their Copilot offer visibility is low despite having active Microsoft Shopping campaigns. An audit reveals that promotional data is added to the feed only at the start of a sale and not removed when the sale ends, causing price mismatches that trigger policy flags. Implementing automated feed updates tied to the promotion schedule resolves the mismatch and restores offer highlight eligibility within two weeks.
Ecommerce example: high-consideration products
A home electronics retailer selling premium audio equipment benefits from Copilot Offer Highlights because their product data is detailed and their comparative specifications are complete. Users asking Copilot to compare headphone options see this retailer's products cited with specific technical attributes alongside the price. Conversion from these AI-cited recommendations is higher than from equivalent Shopping ad clicks because the user arrives having already received a qualified AI recommendation.
Lead generation note: lessons that still transfer
Lead generation advertisers do not have product feeds, but the underlying principle applies: AI systems surface service providers who have clear, specific, structured information about what they offer, who they serve, and what the user can expect. Service pages, pricing pages, and FAQ content all function as the equivalent of product data for AI recommendation of service providers. The same discipline of accuracy, specificity, and structured formatting applies.
A readiness checklist
- Audit GTIN coverage across your catalog — target 95%+ for eligible products
- Verify automated feed refresh frequency — at minimum every 6 hours for price and availability
- Add structured promotions to Merchant Center using the promotions feed format
- Review and complete returns and shipping policy sections in Merchant Center account settings
- Audit product images for resolution, background, and quality standards
- Check Merchant Center account health — resolve any active policy violations before applying for AI commerce features
- Confirm conversion tracking is set up to capture off-site transaction data from AI commerce channels