Case Study · E-commerce & Retail
How GEO Strategy of a European Marketplace Was Changed Based On Detailed AI Search Visibility Analysis
AI Search Monitoring & GEO Strategy for a European Online Marketplace
SERVICES
AEO & GEO Monitoring · AI Recommendation Analysis · Technical & Content Web Site GEO Audit · Review & Reputation Analysis · AEO / GEO Strategy
CLIENT
Online Marketplace
INDUSTRY
E-commerce & Retail
MARKET
Europe
PROGRAMME
Ongoing
Key Findings
5
Monitoring approaches used to understand why the marketplace was or was not recommended
85% / 15%
Initial split between technical and content work and external visibility
Visibility barriers
Product quality, customer service, returns and buyer trust repeatedly appeared in the reasons behind marketplace recommendations
Strategy rebalanced
Reviews, external reputation and PR became a larger part of the GEO work
The Challenge
The client is a European online marketplace with a large product catalogue and a strong position in the low-price segment.
The project started with a technical and content review of the marketplace, covering crawler access, product and category content, how products were assigned to categories, use cases, customer feedback and the information available to AI crawlers. AI search monitoring ran in parallel to understand where the marketplace appeared in shopping recommendations and where other marketplaces were preferred.
When the marketplace was not being recommended consistently, our team changed the monitoring approach and prompt structure several times to understand why. The analysis showed that the main barriers were not coming from the marketplace's own website or product information. Instead, AI systems were repeatedly picking up concerns from external reviews and third-party sources, particularly around product quality, customer service, returns, delivery and buyer trust.
This finding changed the priorities of the GEO work. While technical and content improvements to the marketplace remained important, reviews, external reputation and third-party visibility became a much larger part of the strategy, because these external signals were directly influencing whether the marketplace appeared in AI recommendations.
The Baseline
The initial strategy focused largely on the marketplace itself.
The analysis covered:
- AI crawler access and content visibility.
- Product and category information.
- Product-to-category relationships.
- Product use cases.
- Customer feedback and reviews available on the site.
- Marketplace mentions and recommendations.
- Sources used in shopping answers.
- Reasons given for recommending or not recommending a marketplace.
Based on the website analysis, the first GEO strategy developed in house by the customer placed around 85% of the focus on technical and content improvements and 15% on external visibility.
The technical audit confirmed there was work to do on the site, including content that depended on JavaScript, limited extractable content on some pages and gaps in product structured data.
But the AI search monitoring was showing a different and more significant part of the problem. As our team changed the prompts and tested five different approaches, the same concerns around product quality, customer service, returns, delivery and buyer trust kept appearing in marketplace recommendations.
This showed that the original 85% / 15% split no longer matched what was affecting the recommendations, and the strategy needed to give more weight to reviews, external reputation and PR.
What Was Done
1
Audited the marketplace website and content
Product and category pages were reviewed for crawler access, content availability, category descriptions, product information, product-to-category relationships, use cases, customer feedback and structured product data. The findings were used to build the first technical and content plan.
2
Ran product-focused AI search monitoring
The first monitoring tested product recommendations and the shops or websites suggested for buying those products. This established where the marketplace appeared and where other sellers were being selected instead.
3
Changed the monitoring approach
Our team changed the prompts across four further monitoring rounds to isolate different parts of the buying decision. The monitoring moved from product and shop queries to marketplace-specific questions, category-level marketplace recommendations, low-price shopping queries and direct comparisons between named marketplaces. In total, five monitoring approaches were used to understand how the recommendation changed depending on the way the buying question was asked.
4
Analysed which marketplaces were recommended and why
The answers were reviewed not only for whether the marketplace appeared, but for the reasons given for recommending or avoiding it. Across different prompt types, the same themes kept returning: low prices were recognised as a strength, while product quality, customer service, returns, delivery and buyer trust worked against the marketplace in many recommendations.
5
Analysed reviews and external sources
Reviews and other external sources across five external platforms were analysed to understand what information about the marketplace was available outside its own website and which themes were being repeated in AI answers. The same concerns seen in the monitoring were also present in the external information around the marketplace.
6
Reworked the AEO / GEO strategy
Our team compared the monitoring findings with the original technical and content plan and changed the balance of the GEO strategy. The original 85% / 15% split no longer reflected where the work was needed. Technical and content improvements remained part of the strategy, but more weight was moved toward reviews, external reputation and PR.
7
Continued AI search monitoring
Monitoring continues alongside implementation to track changes in how the marketplace is described and recommended.
HIGH LEVEL OUTLOOK
Why the Marketplace Was Losing Recommendations
The monitoring across 5 rounds showed that the problem was not limited to whether AI systems could access and understand the marketplace website. The marketplace's low-price positioning was already understood, but the pattern changed when the models had to decide whether they should recommend it.
Across five monitoring approaches, product quality, customer service, returns, delivery and buyer trust repeatedly appeared in the reasoning behind marketplace recommendations. Price-focused queries made the marketplace more relevant but did not remove those concerns. Direct marketplace comparisons showed the same pattern, with the marketplace remaining an option for buyers focused on price while other options were preferred when reliability, support or lower purchase risk mattered more.
Based on these findings, our team concluded that improving the marketplace website alone would not address the full recommendation problem.
How the GEO Strategy Changed
The original strategy was weighted heavily toward the marketplace website. Our team's analysis changed that balance towards presence on external resources and sourcing more reviews from existing customers.
Technical and owned content. Continue improving crawler access, product and category information, product relationships, use cases and customer feedback available on marketplace pages.
Reviews and reputation. Make existing positive and balanced customer feedback more visible across the external sources used when marketplaces are assessed.
External presence. Build stronger PR and editorial coverage around the marketplace, its products and customer experience.
Monitoring. Continue tracking the same recommendation scenarios while the strategy is implemented.
Technical and content work remained necessary, but reviews, reputation and external visibility became a larger part of the programme.
Current Focus
The current work puts more weight on what is visible about the marketplace outside its own website. Existing positive and balanced customer experiences are being made more visible across review platforms and other external sources, while PR work is adding more current coverage of the marketplace, its products and customer experience.
Technical and content improvements continue in parallel, and AI search monitoring tracks how the marketplace is described and recommended as the work progresses.
This connected the recommendation patterns seen in monitoring with the marketplace's wider reputation.
In marketing since 2009. In AI search since it became a category.
Our team has worked with international brands across SaaS, e-commerce, fintech and professional services for over fifteen years, originally as ContActive Tech Communications.
Today, that same senior team applies its strategic and technical expertise to answer engine optimisation, generative engine optimisation, and AI search visibility.
If AEO and GEO are on your radar, we would be glad to talk.