AI WhatsApp Storefront: 15% CVR Lift by Meeting Customers Where They Are
The Problem
We had a standard web storefront for merchant acquisition. Conversion rates were flat. Merchants in MENA are deeply WhatsApp-native — it's how they communicate, shop, and do business. Asking them to navigate a web UI was friction we were introducing, not reducing.
Every percentage point of CVR lost at acquisition is compounded downstream. The web storefront had a 60%+ drop-off rate between first product view and completed registration. The hypothesis: it wasn't the offer, it was the channel.
WhatsApp penetration in Saudi Arabia and Egypt is >90% in our target demographic. Building a conversion channel there wasn't a nice-to-have — it was going where the market actually was.
My Role
Product Owner. Owned the full channel: research, hypothesis formation, PRD, WhatsApp flow architecture, AI conversation design, integration with our product catalog and CRM, A/B test design, and launch.
Key Decisions
The temptation was to build a simple 'send catalog' bot. I pushed for a full AI-driven storefront experience: product discovery, Q&A, cart building, checkout, and order confirmation — all in WhatsApp. The complexity was worth it for the conversion impact.
Used the entry point data (which Meta ad they clicked, which product category) to personalize the first WhatsApp message. This increased qualified engagement dramatically — felt like a direct, relevant conversation rather than a generic blast.
Scripted bots break when customers ask something unexpected. The AI needed to handle natural product questions, comparisons, and objections. Invested heavily in training and testing the AI on edge cases.
Experimentation
A/B tested web storefront vs. WhatsApp storefront entry points from the same Meta ads. Measured CVR at each funnel stage: click → conversation → product view → cart → order. WhatsApp won at every stage.
Launch
Launched alongside existing web flow (not replacing it). Within 4 weeks, 60% of new merchant acquisition was coming through the WhatsApp channel organically.
What I Learned
“Distribution is as important as product quality. A 10x better product delivered through the wrong channel will lose to a decent product in the right channel. This project rewired how I think about go-to-market.”