AI recommendations for Shoptet
Boost your store's revenue with AI recommendations
Personalized recommendations powered by Google Cloud AI.
No setup, no commitment.
Simple as 1-2-3-4
How it works
Install
One click in the Shoptet marketplace. No coding required.
Auto-configuration
Automatic catalog sync and widget placement.
AI learns
A neural network analyzes customer behavior in real time.
Revenue grows
Average conversion lift of 3-8% within 30 days.
Intelligent widgets
5 types of AI recommendations
Each widget targets a different moment in the shopping journey
Similar Products (Similar)
"Products with similar attributes to the one you're viewing"
You may also like (OYML)
"Based on your history and relevance to the product you're viewing"
Recommended for you (RFY)
"Based on your browsing and purchase history"
Frequently bought together (FBT)
"Products frequently purchased together in the same session"
On sale (On Sale)
"Discounted products you might be interested in"
10,000 free credits to get started
No commitment. Then monthly plans from 390 CZK.
Interactive preview
Try it out
See how the widgets look with real products
ROI calculator
Calculate your return on investment
Enter your numbers and see the estimated impact
Importing user events from GA4 eliminates the cold-start phase. Vertex AI can immediately activate OYML, RFY, and FBT models instead of just the basic Similar.
Recommended plan
To reach these goals, you need at least — credits / month
Estimated gross profit increase / month
Assumptions table
| Parameter | Conservative | Realistic | Optimistic |
|---|---|---|---|
| Implementation phase | Cold start | 3-6 months | 6+ months |
| Active models | Similar | OYML + FBT | OYML + FBT + RFY |
| Rec click rate (% of visitors) | 3 % | 7 % | 12 % |
| Conversion rate (clicker → order) | 4.5 % | 6.0 % | 9.8 % |
| Incrementality | 10 % | 20 % | 30 % |
| Global APR growth (FBT) | 1 % | 2 % | 3 % |
Sources: Salesforce Commerce Cloud (rec click rate 3-12%) · Dynamic Yield (session conv rate 4.5-9.8%) · Nosto case studies (global APR growth 1-3%) · Vertex AI documentation
Data sources
- Salesforce Commerce Cloud (1,400+ stores): rec click rate 3-12% depending on personalization phase
- Dynamic Yield (400+ enterprise clients): session conversion rate 4.5-9.8% for recommendation clickers
- Nosto case studies: global APR growth 1-3% with FBT widget implementations
- Vertex AI documentation: models with 12+ months of data capture seasonal patterns
What is rec click rate
Rec click rate = the percentage of store visitors who click on at least one widget recommendation. This is not the CTR of a single widget, but the share of all sessions. The numbers in the calculator (3-12%) correspond to industry averages by personalization phase (Salesforce Commerce Cloud).
What is session conv rate
Conversion rate = the likelihood that a visitor who clicked on a recommendation makes a purchase in that session. Typically 2-3x higher than the overall store conversion rate, because recommendation clickers are more engaged (Dynamic Yield benchmark).
What historical GA4 data actually changes
Without GA4 import: Vertex AI starts from scratch. Similar works immediately (only needs the catalog). OYML and RFY need ~1,000 detail-page-view events (4-8 weeks). FBT needs ~1,000 purchase-complete events (2-6 months).
With GA4 import (3+ months): OYML, RFY, and FBT activate on installation day. Rec click rate boost of +20-50% and session conv rate boost of +10-30% correspond to the difference between Similar and fully trained models.
Incrementality — why we don't count all conversions
Most people who click on a recommended product already came to the store intending to buy. The widget just made their path easier — they would have found the product through search or category navigation. The real incrementality of recommendation tools is between 10-30%. The rest is cannibalization of existing conversions. That's why the calculator uses conservative estimates: 10% (conservative), 20% (realistic), 30% (optimistic).
The FBT widget can increase cart value by 10-15%, but only for customers who actually add the suggested item. The overall APR increase across the entire store from a single widget is typically 1-3% in practice. That's why the calculator uses global APR uplift of: 1% (conservative), 2% (realistic), 3% (optimistic).
What the calculator intentionally excludes
- Reduced bounce rate (measurable but hard to isolate)
- Impact on email retention and repeat purchases
- Negative effects of poorly configured recommendations
Sample calculation (Realistic scenario)
Results are approximate estimates based on aggregated industry data.
Real-time metrics
Clear dashboard
Track the performance of Recofy recommendations in real time
Pricing
Simple pricing
Monthly subscription. 1 credit = 1 widget view.
Start
For smaller stores
Growth
For growing stores
Pro
For mid-size stores
Scale
For large stores
Enterprise
Custom pricing
Questions and answers
Frequently Asked Questions
Shoptet offers basic "related products" based on manual assignment or category. Recofy uses Google Cloud Retail AI — a neural network that learns from real visitor behavior and personalizes recommendations for each customer in real time.
Each widget view uses 1 credit, regardless of type. Monthly plans start at 390 Kč (10,000 credits). You get 10,000 free credits upon installation to try it out.
Widgets simply stop showing — the panel disappears gracefully, no error messages, no broken pages. Auto-recharge (490 Kč / 10,000 credits) ensures uninterrupted operation.
No. The widget is only 2.7 KB (gzip), loads asynchronously, and never blocks page rendering. The API has an 800 ms timeout — if the server doesn't respond in time, the panel simply doesn't appear.
Yes. All data is stored in the EU (Google Cloud eu-west1, Supabase EU). Recofy respects cookie consent — the tracker only activates after analytics consent is granted via Shoptet's cookie bar. We do not collect any personal data.
At least 100 active products. The AI needs enough data to find patterns. The Similar Products widget works immediately, while personalized models activate gradually as event volume grows.
The Similar Products widget works within 24 hours. Personalized recommendations improve as interactions grow — typically after 1,000+ views. Full revenue impact within 30 days.
Yes. 10,000 free credits upon installation. No credit card required. After they're used up, choose a monthly plan starting at 390 Kč.
Ready to boost your revenue?
Install Recofy from the Shoptet marketplace and get 10,000 free credits. No credit card, no commitments.
Install from marketplace