Your reviews already contain a complete menu audit. Customers tell you exactly which dishes they ordered, which ones impressed them, and which ones fell flat. The problem is that most restaurant owners read reviews one at a time, react to individual complaints, and never aggregate the dish-level data sitting in plain sight.
Dish names in reviews analytics means systematically tracking every menu item mentioned across all your reviews—positive and negative—to identify patterns you can’t see by reading reviews manually. It’s the difference between knowing that “the pasta was good” and knowing that your cacio e pepe appears in 23% of five-star reviews while your carbonara shows up in 31% of three-star reviews with the word “bland.”
This isn’t about sentiment analysis or star ratings. It’s about extracting the actual dish names customers write, counting them, and matching them to the sentiment of the sentence they appear in. When done consistently, it becomes a zero-cost menu performance tracker that updates itself every time someone leaves a review.
Why Dish Mentions Matter More Than Overall Ratings
A restaurant with a 4.3-star average might assume everything is fine. But if you parse the reviews by dish, you often find a different story: three menu items drive almost all the five-star mentions, two dishes generate most of the complaints, and the rest of the menu is invisible—nobody writes about it at all.
Dish-level analytics surface three categories of menu performance:
- Hero dishes: Items that appear frequently in positive reviews, often with descriptive language (“best,” “perfect,” “amazing”). These are your retention drivers and word-of-mouth fuel.
- Problem dishes: Items mentioned in negative or mixed reviews, often tied to specific complaints about temperature, portion size, seasoning, or value.
- Ghost dishes: Menu items that almost never appear in reviews. They’re ordered, but they don’t inspire customers to write. That’s a yellow flag—either the dish is forgettable or it’s under-ordered.
The monthly Get Kandid Report tracks dish mentions automatically. It reads your reviews every day, extracts the menu items customers name, and groups them by sentiment and frequency. You see which dishes are pulling their weight and which ones are costing you margin without delivering memorable experiences.
How to Read Dish-Mention Data
Raw mention counts don’t tell the whole story. A dish mentioned 40 times sounds popular—until you realize 35 of those mentions are complaints. Context and ratio matter.
Mention Frequency vs. Sentiment
High-frequency mentions in positive reviews are your menu anchors. If your burger appears in 40% of five-star reviews and customers use words like “juicy,” “perfectly cooked,” or “best in the neighborhood,” that’s a signal to protect the recipe, train every cook on it, and feature it more prominently.
High-frequency mentions in negative reviews are operational red flags. If your fish tacos show up in 18% of three-star reviews with complaints about temperature, dryness, or bland seasoning, you have a consistency problem—either in prep, timing, or ingredient quality.
Low-frequency mentions are harder to interpret. A dish mentioned twice in 200 reviews might be a niche favorite or a menu item that few customers order. If it’s labor-intensive or uses expensive ingredients, it might not justify the menu real estate.
Sentiment Clustering Around Specific Words
It’s not enough to know that a dish is mentioned. You need to know how it’s mentioned. Positive clusters often include words like “flavorful,” “generous,” “tender,” “fresh,” or “worth it.” Negative clusters include “cold,” “overcooked,” “small,” “greasy,” “bland,” or “overpriced.”
When the same dish appears with conflicting sentiment—some reviewers love it, others hate it—that usually points to inconsistency. Either different cooks are plating it differently, or ingredient quality varies by supplier batch.
Comparing Dish Performance Over Time
Dish analytics become more valuable when you track them month over month. A hero dish that suddenly drops in positive mentions—or starts appearing in complaints—often signals a recipe drift, a supplier change, or a new cook who wasn’t trained properly.
Similarly, a problem dish that you’ve re-engineered should show improved sentiment in subsequent reviews. If it doesn’t, the fix didn’t work, and you need to revisit it or cut it from the menu.
Using Dish Data to Guide Menu Changes
Menu engineering typically relies on sales mix and food cost. Dish-level review analytics add a third dimension: customer enthusiasm. A dish can have acceptable margins and decent sales volume but generate no positive word-of-mouth. That’s a mediocre menu item—it’s not hurting you, but it’s not helping either.
Promote Your Hero Dishes
If a dish consistently appears in five-star reviews, make it easier for new customers to order it. Feature it on your website, mention it in your Google Business Profile description, and train your front-of-house staff to recommend it. Some restaurants add a “guest favorites” section to the menu and populate it based on review mentions, not just sales data.
Fix or Cut Problem Dishes
When a dish generates frequent complaints, you have two options: re-engineer it or remove it. Re-engineering means changing the recipe, the plating, the portion size, or the price. Removal means accepting that the dish isn’t working and using the menu space for something else.
In our experience, restaurants commonly delay both decisions too long. They tweak a dish once, assume it’s fixed, and never check whether the review sentiment actually improved. Or they leave a low-performer on the menu for years because “some customers like it,” even though those customers never write reviews and the dish quietly drags down the overall experience.
Test New Dishes With Review Feedback
When you introduce a new menu item, dish-mention tracking tells you whether it’s landing. If it starts appearing in positive reviews within the first month, you have a potential hero. If it’s mentioned rarely or appears in lukewarm feedback, you know quickly that it’s not resonating.
This is faster and cheaper than waiting for sales data to stabilize. Sales volume can be influenced by menu placement, server recommendations, and pricing. Review mentions reflect genuine customer enthusiasm—or the lack of it.
Dish Analytics vs. Survey Tools
Some restaurants use post-meal surveys or QR-code feedback forms to gather dish-level data. These tools work, but they introduce selection bias: only a small subset of customers fills them out, and the feedback skews toward extremes—very happy or very unhappy.
Review-based dish analytics capture a broader, more representative sample. Customers write reviews voluntarily, without prompting, and they mention dishes in the context of their overall experience. That context matters. A dish mentioned in a five-star review carries different weight than the same dish mentioned in a three-star review alongside complaints about service or noise.
Review analytics also don’t require any customer-facing changes. You don’t need to print QR codes, train staff to hand out cards, or manage a separate feedback platform. The data is already being created; you just need to organize it.
What the Report Shows
The monthly Get Kandid Report includes a section that lists every dish mentioned in your reviews during the reporting period, sorted by frequency and tagged with sentiment indicators. You see:
- Which dishes appeared most often in five-star reviews
- Which dishes appeared most often in one-, two-, or three-star reviews
- Which menu items weren’t mentioned at all
- Month-over-month trends for individual dishes
This isn’t a real-time dashboard—it’s a monthly summary delivered as a report. That cadence matches how most independent restaurants actually make menu decisions: not daily or weekly, but during quarterly menu reviews or when a dish is clearly underperforming.
The Starter plan ($29/month, or $23/month paid annually) includes dish-mention tracking for one location. The Pro plan ($59/month, $47 annually) adds the Competitor Report, which shows dish mentions from competitors’ reviews—useful for spotting menu gaps or seeing which of your dishes customers compare favorably to competitors. The Premium plan ($99/month, $79 annually) supports up to five locations with consolidated reporting.
You can request the first report free—no card required, no sales call—to see exactly how your menu is performing in customer feedback before committing to a subscription.
Limitations and Hedge Cases
Dish-mention analytics work best for restaurants with a defined menu and consistent dish names. If you rotate specials frequently or use vague menu descriptions, customers may describe dishes generically (“the fish,” “the pasta”) rather than by name, which makes aggregation harder.
The method also depends on customers actually naming dishes in their reviews. In our validation set of 10 NYC businesses with 14,000+ reviews, roughly 60–70% of restaurant reviews mentioned at least one specific dish. That’s a strong sample, but it means 30–40% of reviews contribute only to overall sentiment, not to dish-level insights.
Finally, dish analytics can’t tell you why a dish is underperforming—only that it is. A dish mentioned negatively might have a recipe problem, a plating problem, a price-perception problem, or a training problem. The data points you toward the issue; you still need to investigate and fix it.
Integrating Dish Data With Other Metrics
Dish-mention analytics are most powerful when combined with other review intelligence. For example:
- If a dish appears frequently in positive reviews but your overall star distribution is skewing negative, the problem is likely service, ambiance, or value—not the food itself.
- If sentiment trends show a gradual decline and your hero dishes start appearing less often in positive reviews, that’s an early warning of recipe drift or ingredient quality issues.
- If your review volume is low, dish-mention data becomes noisier and takes longer to yield actionable patterns. Increasing your review volume—by asking customers at the right moment or improving the experience—makes the analytics more reliable.
FAQ
Can I track dish mentions manually by reading reviews?
Yes, and many operators do. The challenge is consistency and time. Manually tracking dish mentions across hundreds of reviews, tagging sentiment, and comparing month-over-month trends takes hours. Automated review monitoring handles the parsing and aggregation so you can focus on decision-making, not data entry.
What if customers use different names for the same dish?
Variations happen—customers write “cacio e pepe,” “cacio pepe,” “the cheese and pepper pasta,” or just “the pasta.” The Report uses pattern matching to group similar mentions, but it’s not perfect. If you notice splits in the data, you can manually consolidate them when interpreting the results. The bigger the review volume, the less individual naming quirks matter.
Does this work for restaurants with large menus?
Yes, but the analysis becomes more complex. Restaurants with 50+ menu items will see more “long tail” dishes—items mentioned once or twice. Focus on the top 10–15 most-mentioned dishes in each sentiment category. Those are the menu items driving most of your review-based reputation, positive or negative.
Can I use dish analytics to test price changes?
Indirectly. If you raise the price of a hero dish and it starts appearing in negative reviews with words like “overpriced,” “not worth it,” or “used to be better,” that’s a signal. Conversely, if a dish remains a hero after a modest price increase, customers are still finding value. Dish analytics won’t tell you the optimal price point, but they’ll tell you when a price change is damaging sentiment.
Conclusion
Dish names in reviews analytics turn your existing review stream into a continuous menu audit. You don’t need surveys, focus groups, or expensive consulting—just a systematic way to read, parse, and aggregate what customers are already telling you.
The monthly Get Kandid Report does that parsing for you. It tracks dish mentions, tags sentiment, and shows month-over-month trends so you can promote hero dishes, fix problem dishes, and make evidence-based menu decisions. If you want to see how your menu is actually performing in customer feedback, start by requesting the free sample report. No card, no call—just the data.