Context & problem
Six steps in,
still no idea what TheFork does.
The existing onboarding puts a new user through six screens (splash, welcome, a three-point "how it works," a tracking pre-permission screen, the ATT system prompt, and a login wall) before they ever see a restaurant or understand what the product is for. Location, the one input that makes the whole value proposition possible ("restaurants near you"), is asked for late and without explanation. The result: users spend the most effort of the entire journey before the app has proven anything to them.
Home & Location
We do not ask for the location, so the first screen already gives us an empty state: negative for the user, who made several steps and finds nothing.
Location's permission
Then, after several steps, we ask for the user's location so we can offer restaurants near them and tell them the value proposition, well after they needed to hear it.
Date & Hour
We are asking a lot of effort from the user, who does not know us yet.
Restaurant listing
After 6 steps, we finally show our listing of different restaurants, based on location, date and hour.
The current flow, screen by screen
Step 6
Login wall
Home
Empty state, no location yet
Permission
Asked late, no context given
Effort
Date & hour, asked upfront
Finally
The listing, after all of that
Mission
A first impression strong enough to keep the user, without demanding six steps of faith first.
We don't ask for location, even though it's the key to the one promise we make: restaurants near you. We don't make the path easy or attractive, and we demand a lot of effort from someone who doesn't know us yet. Those pains push the user toward leaving the app, and toward churn, before they've had a reason to stay.
6
steps before any restaurant is shown
Step 3
the only place the product explains itself
0
personalization before the listing
Wireframes
Three steps, validated in wireframes first.
Before locking any visual direction, low-fidelity wireframes validated the three-step structure itself: one question per screen, a single primary action, and a visible way out (Skip). Content hierarchy and flow were tested here, before cuisine icons, copy tone or the match-score card design were decided.
Measurement
This proposal didn't launch, so there are no results to report. These are the three metrics I'd track, and what I'd expect each one to do.
Onboarding completion
Share of new users who reach a restaurant listing.
Three steps and a visible Skip turn hard drop-offs into soft skips, so more people reach the listing.
Location permission granted
Share of users who allow location.
Explaining the value before the system prompt ("find the best spots for you") earns more yeses than a cold OS dialog.
Conversion to reservation
Bookings that start from the first screen after onboarding.
A Shortlist built from the user's own answers converts better than a generic, unfiltered home feed.
Open questions
What only a launch can answer.
Which step or copy line drives the most Skip taps? Does "Share Location" beat typing a city?
User personas · three scenarios
Old flowSofía · planning Saturday night
Is asked to log in before she knows what the app does. She closes it.
New flowSofía · planning Saturday night
Two taps in, the app already has her area and the food she wants.
Old flowMarco · privacy-conscious
Gets the location prompt with no reason to say yes, and taps Don't Allow.
New flowMarco · privacy-conscious
Reads why location matters before the prompt appears, and is more likely to share it.
Old flowJulia · in a hurry
Goes through six screens before seeing a single restaurant, and gives up on the way.
New flowJulia · in a hurry
Taps Skip and lands on real results, without being pushed to fill everything in.
