Anonymised examples from past projects
Client details are anonymised by request. The figures shown reflect what was measurable in each project. We don't include outcomes we can't attribute clearly to the AI implementation.
All case study data is based on metrics reported by clients or observed in agreed monitoring setups. Results vary between projects based on store size, product type, platform configuration, and team processes. These examples should not be taken as typical or guaranteed outcomes.
Home goods retailer, 8,000+ SKUs
Germany · Founded 2017 · 6-person team
A mid-size home goods store receiving 120–180 support tickets per week. The majority were order status inquiries and return requests — queries that required looking up order data the team already had access to.
Integration: Shopify Plus + Gorgias helpdesk. The agent accesses order data in real time and handles escalation to human agents for complaints and complex returns.
Figures based on 90-day post-launch monitoring period. Client reported no measurable increase in negative support reviews during the period.
Sports equipment store, 450 SKUs
Germany · Founded 2019 · 4-person team
A niche sports equipment retailer with a high cart abandonment rate and a small team that didn't have capacity to run manual follow-up sequences. Previous Klaviyo setup was a single generic email 24h after abandonment.
Integration: WooCommerce + Klaviyo. Recovery sequence varies based on cart value, whether items are seasonal, and customer's previous purchase history. Setup took 3 weeks.
Client chose not to share specific revenue figures. Outcome description based on configuration detail and client qualitative feedback.
Pet care products store, 1,200+ SKUs
Germany · Founded 2020 · 3-person team
A pet care store with a varied catalogue across multiple species and life stages. The default Shopify recommendations were surfacing irrelevant cross-species products. The team wanted recommendations that matched the customer's known pet profile.
Built using purchase history to infer pet ownership. Recommendation logic updated quarterly as seasonal products change. Setup: 4 weeks.
Client declined to share click-through or conversion data publicly. Technical description is accurate.
Kitchenware & cookware store, 3,400 SKUs
Germany · Founded 2016 · 9-person team
A more established store that wanted to roll out multiple agents as a coordinated system rather than separate tools. Priority was customer support and post-purchase workflows. Cart recovery and recommendations were added in phase two.
The support agent references DHL tracking data directly. Post-purchase sequence handles review requests, care guides (for specific product categories), and reorder reminders for consumables.
This is the most complex deployment we've done to date. Ongoing maintenance retainer active since launch.
Want to discuss your specific situation?
The discovery call helps us understand your store before we suggest anything. No generic recommendations.
Book a discovery call