RANK 1AI COLLECTION SORTING
AI sorts your collections.
You grow.
Feature new arrivals and bestsellers get buried; sort by best selling and new products never get seen. Rankine ends that trade-off — reordering every night, with the lift proven by a randomized experiment on your own store.
Launching soon · Works with Shopify · No code in your theme
COLLECTION · FRONT PAGE
AI sorting active
AI analyzes real shopper behavior on your store.
View today's changes
THE MERCHANDISING DILEMMA
Every sort order picks a loser.
THE LOSER: NOBODY
SET UP IN 3 MINUTES
Three steps, then hands off.
01
Connect your store
Secure connection via Shopify.
Connected store
your-store.myshopify.com
✓Connected
02
Turn on AI sorting
We'll optimize while you sleep.
AI sorting
ACTIVE
03
Sit back and grow
We handle the rest.
✓Runs every night
✓You're in control
PROOF, NOT GUESSWORK
Most apps guess.
Rankine proves — then keeps earning.
No blind A/B test you just have to trust. Rankine runs a switchback experiment on your own store and walks each collection through three stages — with a single switch.
ONE SWITCH · AUTOPILOT
3 stages, automatic
01Off
Your original order. Nothing has changed yet.
02Proving
Optimized and original days, paired within the same week.
03Optimizing
The proven lift is locked in — then it runs every night.
You flip it once. Rankine carries each collection from off, to proven, to fully automatic — and surfaces the exact revenue it added.
01A real control
Measured against your store's own pre-install order — not last week, not someone else's average.
02Weekday-proof
Optimized and original days are paired inside the same week, so weekday and seasonality cancel out.
03No p-hacking
The number of test days is fixed up front, from the noise in your data — results can't be stopped early to look good.
04Answered in money
Not a vague “it's working” — the lift is added revenue vs your original order, with a confidence range.
Being honest costs a little: while proving, some days deliberately show your original order. That's the price of a real measurement — and it's over in weeks, not forever.
REAL RESULTS ON YOUR STORE
Not a feeling. A measurement.
Proven by a randomized experiment on your own store — not someone else's average.
+14.2%measured lift
+12.4%orders from new arrivals
+11.3%revenue lift
Illustrative example — results vary by store and category.
ILLUSTRATIVE EXAMPLE
Before (Original order)
After (With Rankine)
New arrivals grow into bestsellers
✓Bestsellers protected
SAFE, SIMPLE, AND UNDER YOUR CONTROL
Safe by design
One-click undo
Revert to your original order anytime.
No code injected
Safe for your theme and your store speed.
Built on Shopify APIs
Official Storefront and Admin APIs.
Change history
See what changed, and when.
SIMPLE, FAIR PRICING
Simple, scalable pricing
Growth
RECOMMENDED
PROOF GUARANTEE
$49/month
Up to 25 collections
Proven revenue lift (A/B)
Request early access
Proven revenue lift (A/B)
14-day free trial · No credit card required · Cancel anytime
PROOF GUARANTEE
Growth and above: if we can't prove your lift within 60 days, we refund what you paid. Inconclusive? Your free period extends instead.
Guarantee details
FAQ
Questions, answered
Shopify's built-in sorts are static rules — best selling, newest, price. Rankine runs a nightly optimization that balances proven sellers with promising new arrivals, and measures the result with a randomized experiment on your own store.
Yes. One click restores your original order at any time, and the change history shows exactly what moved, and when.
No. Rankine reorders via official Shopify APIs and injects no code into your theme, so your storefront speed is untouched.
The first 1–2 weeks are a learning period. Many stores see measurable improvement after that — and the experiment report shows the lift with confidence intervals.
Rankine has a free plan for one collection. Paid plans are $19/month (Starter, 5 collections), $49/month (Growth, 25 collections with A/B experiments), and $149/month (Scale, unlimited collections). Every paid plan includes a 14-day free trial.
On Growth and above, we judge by the switchback experiment during your first 60 days. If the estimated lift is 0% or below with sufficient data, we refund every yen you paid in that period. The decision rule is fixed up front and disclosed in your report — see the guarantee terms for details.
We never count "inconclusive" as "no effect." Instead of a refund, we pause billing and extend your free period until a verdict is possible (up to 60 more days). Until it's proven, you don't pay.
The sorting dilemma
ends tonight.
14-day free trial · No credit card required · Cancel anytime
Bring measured growth to your store — today.
Get in touch
Questions before you start?
Tell us about your store and what you're trying to grow. We usually reply within one business day.
Prefer email? support@rankine.app