UnderusedPost-purchase
Shipping & Delivery Updates
- Update customers at shipped, out for delivery and delivered
- Follow real courier events, not a guessed timeline
- Recommend the next product on the delivered message
Sends onRuns on
Post-Purchase Feedback
A few hours after delivery, one tap to rate the order. Good ratings go to your storefront, poor ones go to a person.
Hi Nour
Your order #4820 has been delivered.
- Classic Tote
How was it?AmazingGoodNot quite right
Type a message
7 MENA brands run this flow on Nonito. Since October 2025 it has reached 30,375 customers, driven 939 orders and attributed 1.2M EGP in revenue.
How we measure this, and what it means for you
Per-brand conversion has ranged from 0.5% to 5.3%.
This flow is run for the reviews and the reply rate, not the revenue. The orders it attributes are a by-product of the thank-you offer.
Real results from brands on Nonito. Yours depend on your product and pricing, so read this as a range, not a forecast. Figures in EGP.
What it is
- A post-purchase feedback flow asks the customer to rate their order a few hours after it arrives.
- The rating is a tap on a WhatsApp button, not a form to fill in or a link to follow.
- Good ratings can be published straight to your storefront through Judge.me.
- Poor ratings go to a person instead, so a bad experience gets a reply rather than a public review.
Who runs it
- Brands whose sales depend on reviews on the product page.
- Restaurants and cafés collecting service feedback, where the score matters per branch.
- Seven brands on Nonito run it today.
- Any brand that wants to know why customers do not come back, from the customers themselves.
How it works
- The order is marked delivered and the flow waits three hours, so the customer has actually opened the parcel.
- A WhatsApp message asks how it was, with three buttons to tap.
- A good answer triggers the review request and, where you want it, a recommendation for what people buy next.
- A poor answer routes into a human inbox with the order attached, and never gets published.
- Whatever they write back is stored against the order.
Why it works
- The customer is already thinking about the order. A week later they are not, and the reply rate collapses.
- One tap is a low enough ask that people who would never write a review will still answer.
- Splitting good from poor before anything publishes is what makes it safe to automate at all.
- Run it for the reviews and the reply rate. The orders it attributes are a by-product, and the conversion range is genuinely low.
Ask Once. Learn Forever. Sell Again.
Just after delivery is the one moment the customer is already thinking about you, and the only moment they will answer a question about the order. Asking then costs nothing and gets a reply. Asking a week later gets silence.
The moment
The order arrives
Your courier marks it delivered and the flow picks it up.
1 of 5
What You Give, What You Get
What you give already exists in your store. What you get comes back automatically.
What you give
- Phone number
- Delivery status
- Order contents
- First nameoptional
- Branchoptional
What you get
- A rating on the order
- Published reviews
- Complaints, privately
- Written feedback
- A score per branch or product
What Makes It Work, And What Breaks It
The difference between recovering orders and annoying people is in these details.
Do this
Ask on the day, not the week
Three hours after delivery. By day three the reply rate falls off a cliff.
Three buttons, no stars
A tap is a decision people will make. A star widget in a browser is one they will not.
Never publish a poor rating automatically
Route it to a person. Publishing it is how an automation turns one bad order into a permanent product page problem.
Watch the inbox before you switch it on
You are inviting complaints. If nobody answers them you have made things worse, not better.
Recommend from real pairings
What people who bought this actually bought next, not what you want to move.
Read it per branch, not just per brand
For anywhere with more than one location, the average hides the problem.
What goes wrong
Firing on shipped instead of delivered
You are asking how it was before it arrived. The replies will tell you so.
A discount attached to the apology
It reads as buying silence. Ask what went wrong first, and offer something after they have answered.
Publishing everything
One automated bad review on a hero product costs more than the whole flow earns.
Judging it on revenue
Conversion on this flow is low by nature. Measure reply rate and review volume, or you will switch off something that is working.
Asking twice
One request per order. A reminder to leave a review is the fastest route to an opt-out.
How It Changes By Industry
Same flow, different timing and wording.
Restaurants & CafésAsk within the hour
Food is judged immediately. Three hours later they have forgotten, or worse, they have moved on.
Fashion & ApparelAsk about fit, not just the itemFit is the reason most fashion returns happen. Asking surfaces it before the return does.
Beauty & FragranceWait for the product to workSkincare cannot be judged in three hours. Push the ask out to a week for anything with a visible result.
Optical & EyewearSplit frames from lensesFrames are judged on arrival. Prescription lenses are judged after a few days of wearing them.
Home, Decor & GiftingAsk the buyer, not the recipientThe person who paid is the one you can message, and often not the one who opened the box.
See how it is builtThe real builder canvas, with the setup direction on every node.
TriggersTrigger Event
- Event
- Order delivered
Delivered, not shipped. Firing early gets you useless answers.
TimingWait
- Duration
- 3 hours
Three hours, so they have opened it.
MessagingWhatsApp
- Template
- review_request
- Buttons
- Amazing · Good · Not quite right
Three buttons. No stars, no link to a form.
LogicFilter
- Condition
- Rating given
This split is what makes the flow safe to automate.
Amazing or Good
ActionsJudge Me
- Action
- Publish review
Publishes to the product page.
RecommendationsProduct Recommendations
- Source
- Bought-together history
Real pairings, not a manual pick.
Not quite right
MessagingResponses
- On reply
- Route to a human inbox
A person, same day. Nothing published, nothing sold.
100%
Before you start
- A delivered status written back to the order by your courier or store.
- A WhatsApp Business sender connected and verified.
- One template with three reply buttons, approved.
- Somebody who will answer a poor rating the same day.
- Judge.me connected, if you want good ratings published to the storefront.
Every node carries its own setup direction. Drag the canvas to move around it, or open it full screen.
Questions About This Flow
Do poor ratings get published?
No. Only good ratings run through the Judge.me node. Everything else routes to a person to resolve first.
Do we need Judge.me?
Only to publish back to the storefront. Without it the flow still collects and stores the rating and whatever the customer writes.
How many people actually reply?
Far more than reply to an emailed review request, because it is one tap on a channel they already have open. Reply rate is the number to watch on this flow.
What results should we expect?
Across seven brands, conversion has ranged from 0.5% to 5.3%. That is low, and it is the wrong measure for this flow. Run it for the reviews and the replies.
Can we ask per branch?
Yes. Tag the branch on the order and the score becomes readable per location, which is usually where the real problem sits.
How long does it take to go live?
It goes live the day your template is approved, provided your courier writes a delivered status back to the order.
Run It Alongside
The flows that protect the order you just recovered.
Second chanceWin-back
Refund Recovery
- Confirm the refund plainly, with timing and no upsell
- Offer store credit toward the next order
- Route quality complaints to a person on your team
See It Running On Your Own Store.
We will connect your store, switch this flow on, and show you what it recovers. Setup takes an afternoon, and you keep everything it builds.