A guest walks into your restaurant, orders two cocktails, loves them, and leaves within the hour. On paper, that looks like a win. Good drinks, happy customer, table turned quickly. But Honey Jain Guha, managing partner at HafaBawa Hospitality Consultants and co-founder of the Hyderabad cocktail bar Oxymorons, sees it differently. To her, that hour-long visit is a clue. And most restaurant owners walk right past it.
On a recent episode of Today’s Special, Honey broke down the one habit she believes separates restaurants that grow from restaurants that guess. It has nothing to do with fancy software or a data science degree. It comes down to a simple distinction most owners never learn to make: the difference between feedback and insight.
What is the difference between feedback and insight in a restaurant?
Feedback is what a customer tells you directly, a Google review, a Zomato rating, a comment on the bill. Insight is what you figure out on your own by digging into why that feedback happened in the first place. Feedback is the symptom. Insight is the diagnosis.
Think of it the way a doctor treats a fever. The fever itself does not tell you much. A blood test does. Feedback works the same way. If a guest says they loved a dish, that is useful, but it does not tell you what to do next. Insight is what you get when you ask why, again and again, until you reach the actual reason behind the reaction.
Honey’s point is that most restaurant owners stop at the fever. They see a good review or a bad one and react to it directly instead of digging deeper. And because customers only tend to leave feedback when an experience is either excellent or terrible, that feedback only ever represents the extremes. The quiet majority, the guests who had an average time and said nothing, are invisible unless you go looking for them.

The clues already sitting inside your restaurant
Long before you need a CRM or a fancy dashboard, Honey says the answers are already sitting in your data and in your dining room. A few examples she shared from her own restaurant:
Guests who order well but leave fast:
At Oxymorons, Honey noticed a pattern of guests ordering two or three cocktails, loving them, and leaving inside an hour. The insight was not that the drinks were wrong. It was that the space was becoming a pre-party stop rather than a destination. The fix had nothing to do with the menu. It was the playlist. Honey’s team raised the tempo of the music after 9.30pm to encourage guests to linger longer, which directly protected revenue per table.

Dishes that never get ordered:
If an item is not selling, the instinct is to assume guests do not like it. Honey’s advice is to check with the serving staff first. In her own restaurant, she found that servers were only confidently recommending two or three dishes out of a full menu, simply because those were the ones they knew how to describe. The unordered dishes were not the problem. Staff training was.
Dishes that keep coming back to the kitchen:
This one is more straightforward. If a dish is being sent back often, there is almost always a genuine issue with the dish itself.
What gets reshared on Instagram:
Honey pointed to a cocktail on her own menu built around a nostalgic flavour combination from her college days. It became one of the most photographed drinks on the menu, not because it was the most complex, but because it told a story. The insight her team pulled from that was not “make more of this exact drink.” It was “storytelling works,” which then shaped how future menu items were developed and described.
None of these required expensive tools. They required someone willing to ask one more question before accepting the first explanation.
The one framework Honey swears by
If there is a single takeaway from the episode, it is this. Whenever something happens in your restaurant, good or bad, ask why. Then ask why again. Keep going until you hit the root cause rather than the surface explanation.
Honey used her own bar as an example:
A couple of cocktails were consistently outselling everything else on the menu. The easy read is to call them best sellers and move on. Instead, she asked why the other drinks were not selling nearly as well. The answer was not that those drinks were worse. It was that servers had defaulted to recommending the same two or three items they knew best, which meant every other cocktail never even got a fair shot with guests. What looked like a menu win was actually a training gap.
This is the habit Honey wants every restaurant owner to build. A daily, five- to ten-minute habit of looking at what moved, what did not, and asking why, until curiosity becomes muscle memory rather than a one-off exercise.

What to actually track every week
For owners without a CRM or loyalty system in place yet, Honey’s advice is refreshingly low-tech. Grab a notebook if you have to. But if you do have access to POS or restaurant analytics software, here is what she recommends watching on a weekly basis:
- Which items sold the most and which barely moved
- Average order value, and how it compares week over week
- How long orders took to reach the table
- Footfall by day, and whether weekday and weekend patterns are shifting
- What time guests are actually arriving
- Which dishes are being sent back, and how often
None of these numbers is complicated to pull. Most restaurant analytics tools already capture them. The differentiator Honey describes is not access to data. It is the discipline of actually sitting with it regularly instead of letting it pile up unread.
Marketing and menu insight are not separate problems
One thing Honey pushed back on is the idea that branding and operations live in separate boxes.
If a bar positions itself as cocktail-forward but keeps attracting guests who only want straight drinks, that is not just an operations issue. It is a sign the brand positioning is not landing the way it is meant to.
Likewise, if guests keep asking servers for recommendations instead of choosing confidently off the menu, that might not be a service problem. It could mean the menu descriptions themselves are not doing their job.
Her rule at Oxymorons has been to protect brand positioning even when it costs short-term revenue, including turning away private events that do not fit the bar’s cocktail-first identity. The reasoning is simple. Every guest who experiences something off-brand carries that impression to the next forty people they tell about it.

Where to start if you have none of this set up yet
You do not need a POS system or a loyalty program to start collecting insight. Honey’s suggestion is to begin with observation. Sit through a shift with a notebook. Watch how guests react when they sit down, when the food arrives, when the bill comes. Note what gets ordered together, what gets sent back, and what people talk about on their way out.
Systems make this easier to scale, but the habit of asking why comes first. Once that habit exists, adding restaurant analytics tools like a restaurant CRM or POS dashboard simply gives you more data to ask why about, faster.

FAQs
Ques. What is the difference between customer feedback and customer insight?
Ans. Feedback is what a customer says directly, such as a review or rating. Insight is the underlying reason behind that feedback, discovered by asking why the feedback happened in the first place.
Ques. Why do most restaurant owners miss insights in their own data?
Ans. Because they treat data collection as a one-time project rather than a daily habit, and because they react to feedback at face value instead of digging into the cause behind it.
Ques. What restaurant metrics should owners track weekly?
Ans. Best and worst selling items, average order value, table turn time, footfall by day, guest arrival patterns, and how often dishes get sent back.
Ques. Do you need a CRM or POS system to start gathering restaurant insights?
Ans. No. Honey Jain Guha recommends starting with direct observation, even with just a notebook, before layering in restaurant analytics software to scale the process.














































































