When I first arrived in Seoul, the shiny capital of South Korea, finding restaurants that I like proved especially difficult. Crowd-sourced reviews were not a thing. It was trial and error, expat Facebook groups and having locals recommend restaurants to me.
Then, at a Softbank Ventures Korea Forum last year, I saw the word ‘MangoPlate’ printed on a paper bag. Someone told me about the young startup. Oh, they’re like Yelp. I didn’t think much of it, until a few months later when I had to bring a few colleagues around. I downloaded the app, connected it to my KakaoTalk account, and checked out what was in the Gangnam area. My colleagues were tired from a work event earlier, so I wisely opted for something in the 100m radius. Tap, tap, tap. Guys! Fried chicken? We loaded the map and hurried over. Now looking back, I think I decided on this particular chicken restaurant because it was really near their hotel and was priced affordably (it has one review and does not have a rating score!).
I quickly forgot about the app again until sometime in May, I had to pick out a homely yet elegant, mid-range restaurant that served wonderful Italian food. It was one of the few times I had to choose a restaurant for a special occasion, which says a lot about how much I hate patronising unfamiliar places, and I wanted to get it right. Under locations, I picked Gangnam, and sorted the restaurants according to the app’s advanced filters: ratings (instead of ‘personalised’ and ‘review count’), Western, and KRW 20,000 to 30,000. I found myself a list of contenders, from a 5.0 brunch star to another rating-topper selling Spanish cuisine.
I worried over the ratings and reviews and weighed the pros and cons of each place in my head. Finally, on the day itself, I peered into the gastronomic wishing well of an app and found myself a 4.8 that seemed promising even with the not-so-perfect grades. Hello, can I make a reservation? Thankfully, it did not disappoint.
Reviews on MangoPlate don’t allow you to give the restaurant a score; instead, it lets users choose between ‘Recommend’, ‘It’s OK’ and ‘Do Not Recommend’, and leave reasons and photos to better explain their choice. Users can also see what reviews other users have left in the past, which prove legitimacy; these are not the owners who log in with their Facebook or KakaoTalk accounts in order to leave a glowing ‘Recommend’ review, extolling the restaurant in all possible ways. These are real people leaving an account of their dining experience. These ordinary folks — not Gordon Ramsey or Martha Stewart — have the ability to influence others to visit a restaurant or deny it any sort of patronage.
I love going through reviews, be it for restaurants, hotels or cosmetics. I raise my eyebrows at 5.0 ratings and sunny reviews left by happy customers, and read everything else with joy. I love and obsess over the 4.8s and 4-and-a-half stars praising the hotel location, decor and receptionist, but complaining about the breakfast being paltry and unsatisfactory, perusing them to no end before I make a decision.
I ignore the 2.0s because there’s just no way I will go for that purchase. Most of the reviews on MangoPlate are in Korean — which means I won’t be able to understand them until I master the language — but all come with either ‘Recommend’, ‘It’s OK’ or ‘Do Not Recommend’.
“We haven’t really marketed to English-speaking communities here as much but you know, it’s already starting to take off (organically) anyway, so that’s a good thing,” says Joon Oh, Co-CEO, MangoPlate, in conversation with e27.
Also Read: MakeMyTrip to invest US$15M in HolidayIQ
It was the second Friday in July, slightly more than a week after MangoPlate announced receiving US$6.1 million from Qualcomm Ventures, YJ Capital and SoftBank Ventures Korea. Oh had just returned from visiting the company’s new office space in Gangnam Finance Centre, literally 10 minutes away from its current office space in Yeoksam-based co-working building Maru 180, which can only accommodate up to 16 employees — the exact number of full-timers MangoPlate currently employs.
MangoPlate: the story
At one glance, MangoPlate is the sort of company that has got everything going for it. Even without a revenue model implemented, it has managed to secure two rounds of funding from venture capital firms. Its four Co-founders all bring different but valuable skill-sets to the table — Oh is the businessman, Michael Roh is the Chief Data Scientist who comes up with the algorithms that personalise recommendations, Hoseok Ryu handles server programming, and Daewoong Kim is the go-to person for client programming. They have been in business for more than two years now, and things are only looking up. Plus, there are no visible competitors.
According to Monica Shim, PR Manager, MangoPlate, the iOS and Android app boasts a million downloads and 250,000 monthly active users. Most of these users are locals from South Korea, while a small number are tourists or expats. All establishments listed on MangoPlate are based in the country.
“This is just an extension of our hobby,” says Oh, speaking for all four Co-founders. Oh knew Roh from college; Kim and Ryu were labmates at KAIST; and Oh knew Ryu from his previous place of employment — Samsung. “So this job is actually not a job, it’s actually for us to have fun.”
The four of them were foodies who wanted to keep track of new restaurants to eat out at. He recalls, “… so why don’t we all put our sources into a spreadsheet? We started tracking all these restaurants on spreadsheets, and then we started writing our tidbits on what to order… what are some quirky things about the restaurant, where it’s located, and we started building this database with name, phone number, address and you know, little pointers, tips, notes and reviews, however you call it.”
Slowly, the “spreadsheet” became a giant database of restaurants. And then their friends wanted in. “We started to get requests from all sorts of people, just random people or friends of friends who wanted that spreadsheet!” says Oh, who estimated the spreadsheet to contain the details of close to 2,000 restaurants. “And it became kinda annoying because at that point, they would tell us things like, “Oh, that one’s closed” or “That one’s not good anymore” or whatever,” he adds.
One day, the four of them met for dinner again, and the spreadsheet came up. Its popularity had made them realise that there could be a way to turn this into an app.
Oh continues, “And we said, “Hey, Danny (Ryu) does server programming, Bernard (Kim) does client programming, Michael … is a data scientist at Apple”, and he is basically the one who does all the smart things behind the scene. He was like, I can actually make this work for each person.”
How does he do that? “Oh, that’s the magic, right? … Taking into consideration your social graph, your preferences, your experiences at different restaurants, the areas you frequently visit. And so on and so forth, right? And taking that into consideration, he layered in this instead of just showing this directory of restaurants like HungryGoWhere or OpenRice, all that stuff is just a directory (where) everyone sees the same thing. Everyone here (MangoPlate) sees everything differently,” Oh says. (Sic)
Everything that makes MangoPlate different
Personalised recommendations. That’s one of the main features often touted by MangoPlate, which helped to set it apart from what Oh called directories or food blogs. Other features include “reliable reviews”, “wanna go’s”, “check-ins, “follow/following” and “picks”. Think Foursquare’s crowdsourced reviews and lists meet Swarm’s check-ins meets Instagram’s follow/following and Popular or Explore page meet Yelp’s restaurant directory.
That dinner conversation made all of them realise that there is a massive opportunity that was just waiting to be seized. “The dominant player actually got shut down by the government. … Naver. WingSpoon. The government shut it down… because the Korean government is really focussing on igniting the startup economy. If Naver has its hands in every little pot, including all the local services — search, fashion, real estate, recipe, I mean, just name it, if they have their hands in all of them, guess what? No one else can actually do it. Because they have the monopoly over search, and when somebody has the monopoly over search, they can use that monopoly and extend … easily. Thanks to Naver, they gracefully shut it down and gave opportunities to (startups) like MangoPlate,” explains Oh, regarding competition.
Other forms of competition have also not done as well as MangoPlate due to the trust factor. “… In the past, some of these services will become very successful and they’ll take money and they’ll start falling because people will lose trust, and once you lose the trust of the consumer, then it’s all downhill,” says Oh.
At the moment, MangoPlate does not collect revenue from users of the app nor does it do so with restaurant owners. All content is generated by those who use its app, and as per Oh, the startup does not manipulate the ranking or reviews of restaurants. Having personalised recommendations as a feature also deter the company from somehow ranking one restaurant above another for revenue since users, while comparing the app with friends, will see the same results and feel duped. Users might not pay money, but they pay the company their trust and time. And if they ever feel cheated, they will leave.
The benefits of a no-revenue model…so far
Oh says that raising money as a startup is always difficult, but MangoPlate’s lack of a revenue model being implemented at the moment doesn’t make it necessarily easier or harder to do so. With the advent of services like KakaoTalk or Facebook, where investors see that businesses like MangoPlate need to start building traffic and acquire customers before generating revenue, raising funding for a company that does not make money has become possible.
“Not just saying, this wasn’t easy. But the thing that I’m saying is that it’s not more difficult than any other startup,” he adds. “Because the investors need to believe in your vision, and it’s your job as a CEO… to convince other people who are going to be investing in your company, because they’re not just investing in the idea, they’re also investing in the team.”
Also Read: Zomato unbundles food delivery into new app
However, just because there isn’t a revenue model or strategy being implemented does not mean that Oh and his Co-founders have not thought about it, or that the company will always stay this way. In the distant future, there might be ways for MangoPlate to generate money from revenue-share models that see it share its platform or resources with food-related services, like delivery. Or the company might go into other industries like beauty.
“I’m hoping that once we dominate this category, we can easily go into beauty, medical, like nails (or) hair, anything that requires a mobile service, and I want to get my curtains installed, I have no idea where and I want them (users) to be able to trust the recommendations and reviews that are coming in!” says Oh.
Most people rely on the dominant search engine Naver and its blogs or their friends to find out where they should get their cars serviced or the next trendy pet cafe, which can be tedious. “When we first started, we were like, “Should we do discovery? Delivery? Reservation and all that.” And we thought we wanted to do all that at once, but discovery in itself is a massive market. So we need to actually dominate that space, and that goes into adjacent markets… but delivery in particular, I would say that we would partner existing players, more or less, just because there’s such great incumbents and I’m very satisfied with their service. In those cases, I would just partner with them and maybe we could have some sort of a deal where we can do revenue-share, right? Delivery, to your question, I would say, probably not, but reservations, I would say we will probably own. Because it’s a very transparent and easy move from where we are into the reservation space,” Oh concludes.
But for now, it’s restaurants and South Korea that has got MangoPlate’s hands full.