BlogMoviesGamesStatsAbout
Back to blog
·4 min read

The Broken 5-Star Food Rating

thought

Hungry, lazy to cook, open GoFood/GrabFood, didn't know what to order, every food looks delicious, almost all of them got 4.5-4.9 rating, I bought one of them, and it was just... "okay... it's edible" although it has a 4.9 rating. But isn't 4.9 supposed to be the best since almost everyone gives 5 stars to the food?? I think the 5-star review system is currently broken in the food industry, especially in Indonesia.

The Problem

Before we get to the point, let's talk about how the 5-star review system is supposed to work in the food industry. 3 stars should be the baseline. It means "the food arrived, it tasted like standard food, and I am not disappointed"—it's the middle of the bell curve. Of course, if the food has higher than expected quality, it should get 4 stars (even 5 stars if it's perfect), and if it's below the expected quality, it should get 2 stars (or even 1 star if it's really that bad).

But in our current ecosystem (especially on GoFood and GrabFood), it doesn't work quite like that:

  • 5 Stars is now the receipt for a successful transaction. It just means "the food arrived, I didn't get food poisoning, the driver was on time"—and also if the food is good, it will get 5 stars too.
  • 1-4 Stars is reserved as punishment for restaurants—for missing items, late drivers, or food that's just bad. TL;DR: 4 stars are bad, 3 stars are terrible, 2 stars are trash, 1 star is hell.

Why does it work like this? Because the platforms penalize merchants and drivers for anything less than a 4.5 average, so we've been socially engineered to give 5 stars out of guilt. We aren't really rating the food objectively; we are rating the transaction too, because our vote can directly affect someone's job.

But the main problem is, although we can see how bad 1-4 stars are, we can't really see how good 5 stars is. This results in: finding great food is really, really hard! Especially if we only rely on the rating.

Ideas

  • Retention Metrics

    While stars are not the most honest review, the decision to make a second order is. Users don't have any reason to order the same food again if it's not good, so if a restaurant has a high retention rate, it means the food is most likely good. 🤤

    Food delivery apps should surface repeat-order rate as one of the main metrics, not just the star rating.

  • Meritocracy Instead of Democracy

    I don't want to sound like I don't like democracy, but if I imagine wanting to order great nasi goreng, I would rather trust the review from a user who often orders nasi goreng rather than random people. So basically, reviews from "foodie" or "expert" users should be weighted more than those from random people. It would be great if the app could tag users as "foodie" or "expert" based on their order history and rating history. A new section could be added to show reviews from foodie or expert users.

  • Comparative Rating

    In my experience, sometimes it's hard to score food objectively using a score system, but if I'm asked to compare two foods, I can easily say which one is better. If the app could implement a feature that lets users compare two foods from previous orders, it would be easier to review and the results would be more honest. But I'm not sure if this approach would have a good effect on merchants, especially since it wouldn't be fair to compare a Rp 10,000 nasi goreng with a Rp 100,000 one. Also, it sounds like winner takes all, which could be bad for merchants that are just starting out. There should be some improvement on this approach though.

Conclusion

I don't have a solution for this problem, but I just wanted to share my thoughts about it. I think it's a common problem for most people who use food delivery apps, and I hope that food delivery companies will consider this problem and come up with a solution.

Share: