PDP conversation starter optimization

This guide explains the PDP conversation starter optimization feature.

What is PDP conversation starter optimization

Conversational agent shows AI-generated conversation starters on your product detail pages (PDPs), so shoppers can ask questions and get instant answers. Each product has a large question pool, but the page shows only a few at a time.

PDP conversation starter optimization enables Conversational agent to choose which questions to show based on shopper behavior. Conversational agent weighs each question's clicks against how often it's shown, then leads with the top performers and updates questions automatically.

How PDP conversation starter optimization helps

Benefits for customers

• Instant relevant answers: Shoppers get pre-purchase questions answered on the PDP, without searching or leaving.
• Faster decisions: Shoppers resolve doubts on the spot and buy with confidence.

Benefits for your business

• More shoppers buy: Conversational agent leads with the most-clicked questions, so more shoppers get answers. This helps improve the engagement rate.
• Less manual work: Conversational agent ranks and orders questions automatically — no manual curation needed.
• PDP stays relevant: Rankings adapt as shopper interest shifts across seasons and new products.

Where optimized starters appear

Surface what shoppers actually ask

On a leather boot page, "How do I care for these leather boots?" earns far more clicks than a generic returns question. Conversational agent moves the relevant question to the front automatically.

Adapt to seasonal shifts

As winter approaches, a waterproofing question starts outperforming others. Conversational agent promotes it during the season without you updating the setup.

Get new products up to speed

When you add products, their questions start on neutral footing and climb or fall as shoppers interact with them.

How PDP conversation starter optimization works

Conversational agent compares clicks to impressions to score each question's performance. High-performing questions gain priority, and weak ones lose it, so the best questions appear more often.

Over time, every PDP leads with the most optimized set of questions that your shoppers engage with the most, so more visitors start a conversation.

Optimization works at the product level, so every variant of a product — each size or color — shares the same optimized set of questions.

Data and learning

Prerequisite: Ensure tracking is active for the actual_show (impression) and trigger_clicked (click) events with component = PDP_EMBEDDED, so Conversational agent can measure engagement.

• Learning needs data. Products and questions with little traffic take longer to optimize, since Conversational agent needs enough impressions and clicks to score them reliably.
• New questions start neutral. Newly added products and questions start neutral until shoppers interact with them.
• Updates on a schedule. Conversational agent refreshes rankings on a recurring schedule, so ranking changes take effect at the next update rather than immediately.

Access conversation starter optimization

To enable PDP conversation starter optimization, contact your Partner Engagement Manager (PEM).

Conversation starter optimization builds on your existing product-page setup in Conversational agent. Your existing question pools are preserved, so turning on optimization refines the questions you already have rather than regenerating them.