Customer stories
What merchants stopped doing once Revlance was running
Case studies from DTC brands that outgrew static merchandising rules. Actual before/after measurements — not projected upside from a sales deck.
Case study — Kartela Home
From 8,000 SKUs and three rules to a performance channel in 48 hours
The challenge
Kartela Home, an Atlanta-based direct-to-consumer furniture brand, had grown its catalog to 8,000 SKUs across furniture, lighting, and home accessories. Their Shopify storefront was running on three static merchandising rules — all written in the same week their first hire joined. The rules never changed because nobody had time to change them, and the VP of Growth couldn't tell whether the product order was helping or hurting conversion. Every shopper saw the same homepage module. Every collection page surfaced the same top-20 products regardless of what a visitor had been hovering on.
How Revlance helped
Kartela connected Revlance via the Shopify app in under 45 minutes. Within 48 hours, the behavioral model had enough session data to begin influencing grid order. Revlance replaced the static homepage module with a real-time personalized feed, and reranked 4 collection pages — furniture, lighting, outdoor, and rugs — based on individual scroll and hover patterns. The team made no manual rule changes. The engine started optimizing on its own.
+19%
conversion rate across personalized grids
-22%
grid bounce rate after first month
<48h
from integration to first personalized grid
We had 8,000 SKUs and three merchandising rules. Revlance took that from a maintenance problem to a performance channel. I stopped thinking about grid order as a task on my to-do list.
Case study — Mondre Apparel
Email recs that showed what shoppers were considering — not what they already bought
The challenge
Mondre Apparel's Klaviyo email flows were recommending products from the same category as a customer's last purchase. If you bought a jacket, you got more jackets. The problem was obvious to everyone on the team but there was no easy fix — Klaviyo's native recommendation logic doesn't factor in session-level browsing behavior. The Head of E-commerce had watched the email CTR plateau at the same number for six months. They were sending more email, not better email.
How Revlance helped
The Revlance Klaviyo integration added a dynamic rec block to three existing flows — the post-purchase sequence, the browse-abandonment flow, and the weekly new-arrivals broadcast. The blocks pull from the behavioral preference model, not from purchase history. A shopper who spent 4 minutes hovering on outerwear before buying a t-shirt gets outerwear in their next email — because that's what their behavior said they were shopping for. Mondre saw results within the second send cycle after enabling the blocks.
+31%
email click-through after enabling rec blocks
+14%
email-attributed revenue in the first 60 days
3
existing Klaviyo flows updated with no copy changes
The hover signal insight alone changed how we think about above-the-fold grid ordering. We shipped 3 grid variants in a week after seeing which categories our shoppers were lingering on versus buying. That was a new way of thinking for us.
More from the merchants using Revlance
We were manually updating our 'recommended for you' section every Monday. Revlance made that irrelevant. The grid updates itself and it's consistently better than our guess.
I was skeptical that behavioral signals at our volume — 80,000 monthly sessions — would produce useful affinity scores. They did. By week three the model was surfacing category preferences I wouldn't have guessed from purchase history alone.
The merchant dashboard is what surprised me most. I can see exactly which signal drove a recommendation. That transparency matters — I can explain the grid order to my CMO without saying 'the algorithm decided.'
Setup was genuinely fast. I'd heard 'easy integration' from every tool vendor, but Revlance was the first time I actually trusted it. Shopify app, 30 minutes, first personalized results that afternoon.