By Kishan Vasani
Co-Founder & CEO
Most people choose what and/or where to eat by reading reviews and ratings of restaurants. And broadly speaking, the system works, right? Well yes, and no.
Think about all the restaurants you’ve visited or ordered food from over the last few months, especially the new places. How did you choose that restaurant? Probably by reading ratings and reviews. Did the experience live up to what the reviews told you? If so, lucky you! But many people, myself included, want a more reliable way to discover great food and restaurants without depending upon information that probably isn’t relevant. Allow me to explain.
Firstly, let’s be clear about exactly what ratings and reviews are. A review is the opinion of a someone’s own experience at a specific point in time. Some of it might be factual, but it can be exaggerated, especially at the extremes of the spectrum. A rating is simply a value given to that experience. Because it’s not scientific or formulated, it can be problematic. For example, what one person rates 3.5 out of 5, another might have given 3 out of 5.
Despite these issues, the reality is that we use ratings and reviews to make decisions in many aspects of our lives… just think about the last time you bought something from Amazon or Flipkart. The chances are that, for purchases of commodities like phones, nappies or photo frames, reviews were quite helpful. But when you’re using them to make food related decisions, they can be useless.
Why? Your taste for food is personal, extremely personal. Everyone has their own preferences, based upon their background and experiences. We even have irrational preferences. Here’s a sample of my taste profile:
– I eat meat but not beef or fish/seafood
– I hate most boiled food
– I can’t handle food that’s spicy
– I love most sweet dishes but not Indian desserts
– I avoid most rice dishes unless it’s egg fried rice at a Chinese restaurant
I could go on and on and on. We’re complex and strange beings. And that’s not the end of it. What you want to eat is often based upon your food mood. One day you feel like something fried, another day you really fancy Thai food. And that’s just the factors related to food.
How you’re feeling and the context around the meal are major influences too. Imagine you just had a crazy day at the office, you didn’t even get time for lunch. Now it’s dinner time and you’re starving but you’re too tired to cook. You decide to order food. But what should you eat? You’re going to want the food to be delivered super fast, and the food better be hot, filling, and make you feel good.
Ratings and reviews don’t help much here. The last thing you want to do spend more time reading and comparing. Also, it really doesn’t make sense to sum up a restaurant or meal with a score. There are so many facets to the restaurant and dining experience… chef, menu, price, ingredients used, delivery time, service, ambience etc. So if a restaurant has a rating of 3.7, what does that really tell you about each part of the experience? Maybe the food is good but the service not so much. Or is it the other way round? Add to this the reality that sometimes ratings can be manipulated, paid for, or out of date, and you’ve got a broken system.
Below are the search results for a high end restaurant in London:
You’ve got a real mixed bag of ratings here, going as low as 1 star, all the way up to 9.2! What exactly am I meant to infer from these results?
I’m sure many of you have experienced a similar situation to the one I’m about to describe. I love dahi puri, I can’t get enough of it! I have it about 3 times a week. I’ve found a place near the office that makes it great. I found this restaurant while driving to work one day. I didn’t bother to check its reviews and rating before going. But one day curiosity got the better of me and so I read the reviews. It was rated 2.9 on Zomato. The opposite is also true. How many times you have been to a highly rated restaurant only to be disappointed in some way?
That’s all before I make my biggest point… ratings and reviews are likely to be irrelevant to the person reading them! They are the aggregated and averaged opinion from a bunch of strangers. The only possible way a review and rating makes sense is when you know how similar the tastes of the reviewer are to yours. Personalised recommendations based upon your taste preferences, mood and context, are a far better way to receive relevant options for you to make an intelligent meal choice.
Irrelevant reviews don’t just come from strangers. My best friend and I don’t have similar food preferences and when we dine out our expectations of the restaurant are different. He wants a great vegetarian selection, and I want a good wine list. He wants fast service and I want a place with a vibe. So even though he’s my best friend, why would I ask him for a food or restaurant recommendation? It just doesn’t make sense.
I’m willing to bet that every single restaurant can make at least one dish that is loved by someone. The whole ecosystem needs a shift in mentality. Restaurants need to focus on their cooking strengths and create a shorter, balanced menu. Diners need to be smarter and not use or be influenced by scores and experiences that aren’t relevant to them.
The true nature of discovery means taking a risk. Sometimes that risk pays off, other times it doesn’t. When we don’t want to take a risk, we eat the known dishes from our favourite restaurants. And while that feels reassuring, equally we crave new food experiences. I encourage you to make your future food choices without using reviews and ratings, and see how the experience compares. You could do this in a couple of ways:
1) Go to or order from a restaurant even it if has bad reviews and pick a dish that you normally wouldn’t eat.
2) If you live in Bangalore and have an Android phone, you could just use the dishq app.
3) Ask for a recommendation from a food blogger or someone with the same taste as you.
None are ideal but that’s what dishq is looking to change. We want the online food industry to replace ratings and reviews with recommender systems based upon personalisation. Until then, bon appetit. 🙂