Case studies
E-commerce · Amazon & eBay Sellers

Giving Amazon and eBay sellers what the marketplace never would: their own customer relationships

When you sell on Amazon, the customer belongs to Amazon. We built MyQRGuide for a marketplace-seller SaaS client: a QR-powered post-sale platform that captures customer data, prevents negative reviews, and drives repeat purchases.

Client: SaaS platform for online marketplace sellers

LaravelMySQLAI ChatbotWhatsApp Business APIEmail delivery
MyQRGuide: Post-Sale Customer Platform architecture diagram
8

Professional mini-site themes

100%

Customer data owned by seller

0

Marketplace policy violations

Negative review rate after deployment

Case study
The challenge

Amazon and eBay sellers face a structural problem: they fulfil the order, but the marketplace owns the customer. No email address. No phone number. No way to follow up, upsell, or resolve a problem before it becomes a 1-star review. Every sale is effectively a one-time transaction, even for repeat buyers.

  • Marketplace owns the customer: sellers have no contact data after the sale
  • Dissatisfied customers go straight to Amazon reviews instead of contacting the seller
  • No upsell or re-engagement possible without customer contact information
  • Post-sale service forced through slow, clunky marketplace messaging systems
  • Every sale is a one-time transaction regardless of customer loyalty
What we built
  • QR code generation: unique per product or SKU, trackable scans, included in packaging
  • Branded mini-sites: 8 professional themes on seller subdomains (sellername.myqrguide.com)
  • AI chatbot (Guidy): answers product questions and handles post-sale queries automatically
  • Seller dashboard: scan analytics, customer database, campaign performance
  • WhatsApp and email marketing tools: upsell and re-engagement to collected customer base
  • Review prevention flow: captures dissatisfied customers privately before they leave public reviews

The branded mini-site architecture deserves specific explanation. An early design considered hosting post-sale content on a central myqrguide.com domain with a seller path (myqrguide.com/seller-name). We rejected this because it would make MyQRGuide the brand the customer remembers, not the seller. Every seller gets their own subdomain (sellername.myqrguide.com) with their own logo, colors, and domain-level identity. The seller owns that relationship. We are infrastructure. The WhatsApp Business API integration was similarly deliberate: in markets like the UK, India, and the Middle East, WhatsApp has higher open rates than email by a significant margin. Restricting to email would have halved the platform's effectiveness for international sellers.

The results

Sellers capture customer contact data they would otherwise never have. Negative review rate drops because dissatisfied customers have a direct channel to the seller before they reach the public review form.

Before MyQRGuide, every customer we sold to on Amazon was effectively anonymous after the sale. Now we have a real customer database we own. The review prevention flow alone paid for the platform. One avoided negative review on a competitive listing is worth far more than the monthly subscription.

MyQRGuide sellerUK-based Amazon marketplace seller
The outcome

MyQRGuide turns a structural marketplace limitation into a competitive advantage. Sellers who use it build a customer database from every single sale. Something Amazon and eBay will never give them. The platform has zero marketplace policy violations by design: the QR code interaction happens after delivery, entirely outside the marketplace transaction.

Building for marketplace sellers taught us that compliance architecture is not a constraint to work around. It is the core product value. Every design decision we made with platform policy in mind ended up being the right call for the seller too. Sellers who want shortcuts around marketplace rules eventually lose their accounts; sellers who operate cleanly compound their advantage over time. The most defensible SaaS products are the ones that align their users' incentives with the platform's rules rather than helping users circumvent them.

View live project
Continue reading
RAG Pipeline for AI Knowledge Assistant
AI Engineering

RAG Pipeline for AI Knowledge Assistant

>80%
Retrieval hit rate (relevant source in top 5)
LaravelPHPMySQL
Read case study
Automated Product Image Pipeline
Luxury Commerce

Automated Product Image Pipeline

5 min
Photo to S3 (was 3 days)
PythonNode.jsAWS S3
Read case study

Start a project

Ready to build something like this?

30-minute discovery call. No pitch deck. We talk through your problem, tell you honestly if we can help, and scope it properly before quoting.