How to Measure Which Dishes Are Most Requested on WhatsApp at Your Restaurant

If you own a pizzeria or any other food business, it surely happens to you: between 12:00 and 15:00 the phone doesn't stop, and the questions are almost always the same: "Do you deliver to such neighborhood?", "How much is the napolitana pizza?", "Do you have a gluten-free option?". In that chaos, it's easy to misplace an order or not notice what people ask about most. But that info is gold: knowing which dishes generate the most inquiries helps you plan the menu, the promos, and even the stock.

The problem: you have the data, but you don't see it

Every message that comes into your WhatsApp is data: what people ask, what they order, what they doubt. The problem is that if you don't record it, it gets lost. Having "an idea" of what sells best isn't enough: to decide well you need numbers. And if you don't reply quickly, the customer will order from somewhere else. So measuring isn't a luxury, it's a necessity to avoid losing sales.

How to measure which dishes are asked about most on WhatsApp, step by step

1. 1Create a simple spreadsheet (Google Sheets or Excel) with columns: date, dish asked about, type of inquiry (price, availability, delivery), and whether it ended in an order.
2. 2Every time you handle a message, record the dish they asked about. If there are many, schedule 10 minutes at the end of the day to enter the data.
3. 3At the end of the week, count how many times each dish was mentioned. Make a ranking: the top 5 most asked about.
4. 4Combine that ranking with your actual sales: if a dish is asked about a lot but sells little, it could be an issue of price, description, or availability.
5. 5Use that info to decide: which dishes to put in the daily promo, which to highlight on the menu, and whether it's worth adding an option that people ask for often.

How Wando makes measuring easier for you

Keeping a manual spreadsheet works, but it's extra work you don't have time for at midday. That's where Wando comes in: the AI suggests replies for messages and your team approves them with one click. But what's important for measuring is that Wando learns from your business's real replies: it records conversations and groups them, so you can see which dishes are mentioned most without having to enter anything manually.

• Wando uses Meta's official WhatsApp API, so your chats stay organized and searchable.
• The message inbox with AI suggestions lets you reply faster, and that prevents the customer from going elsewhere.
• Wando's statistics show you inquiry patterns: what they ask, at what time, and how often.
• It doesn't reply on its own: you approve each reply, so you always have control.

With Wando, measuring is not a separate procedure: it happens on its own while you serve. And since it's designed for SMBs in Latin America, it speaks your language and has local support.

Common measuring mistakes (and how to avoid them)

• Relying only on memory: "I know what they ask for most" is not enough. Data can surprise you.
• Measuring only sales and not inquiries: a dish that is asked about a lot but doesn't sell is an opportunity, not a failure.
• Not recording the time of the inquiry: knowing what time they ask helps you prepare your team and stock.
• Not acting on data: measuring just to gather dust is useless. Use the info to change something, even if it's a promo.

The three cases that deviate from the happy path (and what to do)

Not everything is a clear message like 'how much is the muzza?'. Sometimes the customer sends a 40-second audio, writes to you at 3 AM, or asks about a dish that's not on the menu. These are the cases that break the mold and where the difference between good and bad service is most noticeable.

First: audios. People in the food industry send audios because they're in a hurry. Wando transcribes them, so you don't have to listen to each one to understand what they're asking for. Second: messages outside business hours. Nobody expects you to reply at 3 AM, but the message should be logged and that when you open the place you already have all the context. Third: questions that aren't orders, like 'do you have a vegan menu?' or 'do you host events?'. These aren't answered with a template: they require human judgment.

In all three cases, the rule is the same: AI gives you a suggestion, but you approve it. If you're not sure, don't send it. Better to take two minutes to respond well than to reply with anything and lose the customer.

How to measure whether management improved

Measuring only the number of messages is not enough. A business can receive more inquiries and sell the same. What matters is conversion: how many inquiries end up in an order.

To measure it well, look at three numbers: response time (how long it takes you to answer the first message), conversion rate (how many conversations end in a sale), and average ticket (whether people order more when you respond quickly). With those three, you'll see if management has truly improved.

Wando gives you message and campaign statistics, so you can see the evolution week by week. But the final analysis is yours: if response time dropped and sales went up, you're on the right track. If not, adjust something and measure again.

• Response time: from messages to the first response.
• Conversion rate: inquiries that end up in an order.
• Average ticket: how much each customer who orders via WhatsApp spends.

What to do when the customer doesn't reply

It happens often: the customer asks about the menu, you reply right away, and... silence. They didn't reply. What do you do? First: don't insist with messages. Nobody wants their chat flooded.

Second: wait a reasonable amount of time. If a few hours have passed and they haven't replied, you can send a follow-up message, but with value: 'Do you have any questions about the order? If you want, I'll reserve it for you for noon.' That's not spam; it's a useful nudge.

Third: don't take it personally. Many times the customer was comparing prices with another place or their signal went out. What matters is that your response was quick and clear, because if they come back, they'll prefer the place that replied well. And if they don't come back, let it be their decision, not because you weren't there.

How it was done before and why that failed

Before, handling WhatsApp in a restaurant was chaos. The owner had a personal cell phone, waiters took orders by phone, and inquiries were answered when possible, not when the customer needed it. The result: lost messages, angry customers, and zero data.

The underlying problem was that there was no record. Each inquiry was an anecdote, not data. You didn't know which dish was asked about most, when peak times occurred, or how many sales were lost due to slow responses. Everything was resolved with 'experience,' which is another way of saying 'I don't have numbers.'

And the most serious thing: today's customer doesn't wait. If you don't reply within minutes, they go elsewhere. Word of mouth is no longer enough. So, what used to be 'answering the phone' is now a process that can be measured and improved. And for that, technology helps, but the judgment is still yours.

What NOT to automate in this process

Automating everything is tempting, but it's a mistake. There are things a bot shouldn't do, no matter how smart it is. First: negotiating prices. If a customer asks for a discount, that requires human judgment. AI doesn't know your profit margin or whether it's a good idea to close the sale.

Second: resolving complaints. An angry customer needs a person to listen, not an automatic response. If you automate it, you'll only make them angrier. Third: decisions about the menu or stock. AI gives you data, but the decision to change a dish or launch a promo is yours.

The golden rule: automate the repetitive (greetings, schedules, frequently asked questions) and leave the human touch for what needs judgment. Wando suggests responses, but you approve each one. That way you don't lose control or your customers' trust.