Growth for ecommerce, starting from what buyers ask.
Below is the working prompt set for this vertical — 8 questions your buyers put to an assistant before they ever reach your site. Everything we build is scored against it.
Get your prompt setThe prompt set
8 questions we measure against
- 01Where can I buy sustainable running shoes that ship to the UAE
- 02Which store has this in stock in my size
- 03Compare warranty and returns between these two retailers
- 04Is this retailer legitimate and what do reviews say
- 05Cheapest total cost including duties and shipping
- 06Best alternative to this product for a lower budget
- 07How long does delivery actually take to my country
- 08Store that offers free returns on international orders
Your live set is built around your categories, locations and competitors. This is the starting template for the vertical.
What makes this vertical different to work in
Product pages carry manufacturer copy identical to every competitor
Assistants compare on total landed cost, which is rarely stated anywhere
Review volume decides recommendation and nobody owns the process
What we usually run here
Example implementation framework. What follows describes how we would approach a ecommerce engagement. It is method, not a record of a completed client project, and contains no performance figures.
Business challenges
Manufacturer copy is identical everywhere
Nothing distinguishes the listing, so price becomes the only variable.
Total landed cost is never stated
Assistants compare on it, and almost no store publishes it.
Review volume decides recommendation
And in most stores nobody owns the process that generates reviews.
AI marketing, SEO, GEO and AEO strategy
Original product content
Specification detail, use cases and honest comparison replacing syndicated manufacturer copy.
Total cost clarity
Duties, shipping and returns stated up front — the comparison assistants actually run.
Category answer hubs
Buying guides that answer the selection question rather than listing products.
Systematic review generation
A repeatable post-purchase ask, measured by product line.
30, 60 and 90 day action plan
- Days 1–30
Baseline and foundation
- Prompt set across product and comparison queries
- Baseline recorded
- Product data audit
- Core Web Vitals audit
- Days 31–60
Structure and publish
- Product and Offer schema with full attributes
- Original copy on top categories
- Buying guide hubs
- Total cost content
- Days 61–90
Convert and corroborate
- Checkout conversion work
- Review programme live
- Marketplace alignment
- First re-measure
Expected workflow
- 01DiscoverBuild the prompt set from real buyer language, record the visibility baseline across four assistants, and audit the technical foundation.
- 02DiagnoseRank findings by commercial impact rather than severity label, and agree what qualified means with the people who take the calls.
- 03BuildRemediate the foundation, deploy the schema graph, and restructure templates so a complete answer sits at the top of each page.
- 04PublishShip answer-first content against the highest-value prompts, with every statistic carrying a named source and date.
- 05CorroborateBuild the mentions, listings, reviews and coverage that let an assistant verify claims somewhere other than your own site.
- 06MeasureRe-measure against the recorded baseline monthly and report pipeline first, with attribution limits stated in the report.
Deliverables
- Category prompt set and recorded visibility baseline
- Technical audit with a prioritised, costed fix list
- Schema graph across every page template
- Answer-first content programme
- Citation and corroboration programme
- Conversion instrumentation and attribution model
- Monthly report opening with pipeline created
AI tools and automation
AI visibility
- ChatGPT
- Google Gemini
- Perplexity
- Microsoft Copilot
- Claude
Search & measurement
- Google Search Console
- Bing Webmaster Tools
- GA4
- Looker Studio
Technical
- Screaming Frog
- PageSpeed Insights
- Schema validators
- Lighthouse CI
Automation & CRM
- HubSpot or equivalent CRM
- Workflow automation
- Lead routing and alerting
- Conversational agents
KPIs ecommerce businesses typically monitor
Citation rate on product-selection prompts
The comparison question that precedes purchase.
Revenue per session
Cuts through traffic vanity to the number that matters.
Checkout completion rate
Usually the largest single recoverable leak.
Review coverage by SKU
Directly drives both conversion and recommendation.
Return rate by product line
Rising returns usually indicate a content accuracy problem, not a product one.
Frequently asked questions
Yes — it is identical to every competitor's, so nothing distinguishes you. Original specification detail and genuine comparison content are the fix.
Find out what AI says about you right now.
We run your real buyer prompts through four assistants and send you the transcript, with your position and your competitors’. No charge, no call required to receive it.
Typical turnaround: 3 working days.