I'm an international student who lives between Raleigh and Hyderabad. This is my teardown of Yelp for people like me, plus five things I'd change and how I'd decide which to do first.
When I landed in Raleigh for my M.S. in Computer Science in Fall 2025, Yelp was one of the first apps I downloaded. I didn't know anywhere. I didn't know which restaurants were good, which areas were fine, or who to ask. I figured Yelp would fill that gap.
Mostly it did. Then I went home to Hyderabad and it didn't. That gap is what this teardown is about. I've used the app in both cities, and I checked what I saw against Yelp's app store listings, its public company disclosures and its product docs. For every problem I found, I've tried to suggest a fix and a way to measure whether it worked. A list of complaints isn't much use to a product team on its own.
Yelp is a marketplace with two sides. People use it to find and review local businesses: restaurants, plumbers, dentists, car shops. Businesses use it to get found, take bookings and leads, and keep an eye on their reputation.
The app matters less than what's behind it, which is more than 200 million reviews collected since 2004. Those reviews are why a Yelp result feels more trustworthy than a pin on a map.
Over the years Yelp has folded a bunch of separate jobs into one app: search and filters, reviews and photos, table reservations, delivery and pickup, and requesting quotes from service pros. It makes its money mostly from advertising, plus paid leads for home-service businesses. So consumers use it for free, and the paying customer is the local business buying visibility.
Where it's strong is the US, Canada, the UK and a few European countries. Yelp has tried to grow elsewhere (and pulled back from a number of international markets around 2016) but never got enough reviews in those places for the app to be useful. I think that one fact explains more of what follows than anything else.
Yelp says its audience is anyone looking for a local business. In practice I see three kinds of people.
What I wanted was pretty simple. I wanted to find food and services I could trust without having anyone to ask, and I didn't want to learn a different app every time I flew between the US and India.
Yelp is clearly designed around the first two people. It treats the newcomer as a slightly less experienced local regular, and most of my problems come from that.
I open Yelp, allow location, and search "Indian restaurants near me." I get a list with star ratings, a price level and a few review snippets. I filter by rating and "open now," pick one, and skim the reviews.
That's where I get stuck. The reviews are in English and written by people I know nothing about. Is a 4-star "great food" from someone who's tried a dozen Indian places around the Triangle, or someone's first ever? I can't tell. I go anyway. If it's good I might check in or leave a review later. Haircut, oil change, dentist, it's the same loop: search, filter, skim, decide.
Then I visit home, open the same app out of habit, and search for a restaurant near my parents' place. The results are thin. A few listings, most with hardly any reviews, some with old hours or photos. That's not a glitch. It's just how far Yelp's coverage goes.
So I close it and open Zomato or Google Maps, like nearly everyone there does. My saved places, my review history, my sense of how to read a Yelp rating: none of it comes with me.
When I'm back, Yelp works again. Nothing about the app changed. I did, twice in a few weeks, and Yelp doesn't notice. There's one version of the product, built for one kind of person in one kind of place, and everyone else works around it or stops using it.
In its home markets, Yelp deserves its spot on my home screen.
None of this is a big surprise next to Yelp's competitors. But it's done well, and it's why I still open Yelp first in Raleigh.
Each idea is small enough for one product team to take on. None needs Yelp to reinvent itself, and each comes with a metric so you'd know if it worked.
Go beyond vegetarian and vegan. Add Jain and halal, and regional cuisine tags (South Indian, North Indian, Indo-Chinese, for example). Let restaurants tag themselves, and check those tags against what reviewers say.
Success metric. How many food searchers use the new filters, and whether those searches lead to a visit more often than searches with the old filter.
An opt-in mode. It could switch on when someone searches a lot in a market they've never used Yelp in, or when they tap "I'm new here." It would show starter lists for the area, a short neighborhood guide, and a quick note on how ratings tend to run locally.
Success metric. Seven-day retention and how many people book something in their first days, compared with similar new-market users who don't get the mode.
An optional label on reviews that shows who wrote them, like "regular reviewer in this city" or "visiting." It gives a newcomer the context a local reads instinctively.
Success metric. How often people mark labeled reviews helpful compared with unlabeled ones, and fewer support disputes about reviews.
When Yelp sees someone has changed countries, it should bring their saved lists, review history and rating context with them. And if they've landed somewhere with thin coverage, it should say so instead of quietly showing worse results.
Success metric. Seven-day return rate after a country change, compared with today, when the drop-off is silent.
Yelp shouldn't try another expensive launch in India. A better bet is a data-sharing or co-listing deal with a local platform that's already dominant. Start in cities with lots of returning students, visa holders and NRIs, who already use Yelp somewhere else.
Success metric. Growth in listings and reviews per listing in the pilot cities, and how often diaspora users go from search to action.
I scored each idea on reach, impact, confidence and effort, roughly following RICE. I treat effort as a cost, not a divisor, so the table stays easy to read.
| Proposal | Reach | Impact | Confidence | Effort | Priority |
|---|---|---|---|---|---|
| Granular dietary and cultural filters | High | Medium | High | Low | P0, do first |
| "New to the city" mode | Medium | High | High | Medium | P1 |
| Reviewer context labels | High | Medium | Medium | Medium | P1 |
| Cross-border profile continuity | Low to medium | High | Medium | High | P2 |
| Local data partnerships | Low (near term) | High (long term) | Low | High | P3, strategic bet |
The table hides something the chart makes obvious: none of these five is a low-impact idea. What separates them is how much work they take. Pick one to see where it sits and why.
I placed these by judgment, relative to each other. It isn't measured data.
Better dietary and cultural filters come first. It builds on a feature people already use, needs little beyond a tag list and restaurants tagging themselves, and you can measure it straight away. Onboarding mode and reviewer labels are next. Both go straight at the newcomer problem for a fair amount of work, and neither has to wait for Yelp to solve international coverage. Cross-border continuity is valuable but reaches deeper into accounts and data, so I'd do it after newcomer mode has been proven in one market. The partnership idea matters most to someone like me, because it's the only one that fixes the India side. It's also a multi-quarter bet across several teams, with real risk, so I'd treat it as a strategic project on the roadmap, not something for the next sprint.
Yelp is really good for people who already know their city. It's a lot worse for people who don't yet, or who move between two. That's not a rare case. International students, people who've just moved, and families spread across countries are a big group, and they tend to need answers right now. Fixing this wouldn't turn Yelp into a different company. It would mean giving the newcomer a place in a product that currently only has room for one kind of user.