Key takeaways
- Track a handful of measures that predict return visits: repeat-guest share, review trend, complaint rate, ticket times, no-shows and your hygiene rating.
- Guests check: 55% of consumers had looked up a food business's hygiene rating in the past year, and 52% of those who know the scheme would accept 4 at the lowest[1].
- Read reviews as a trend, not a score. Some reviews are fake: government research estimated 11% to 15% in common product categories[2].
- Compare your trend with the market, not just last month: managed groups' like-for-like sales were up 2.7% in September 2026[3], while many smaller businesses reported falling demand[4].
Why a short list beats a dashboard
Most venues have more numbers than they use: EPOS reports, booking system exports, review platforms, social insights, mystery-guest scores. The risk is not too little data but too many measures that nobody acts on. A useful guest experience scorecard has six to eight measures, each with a clear owner, a source you already have and a point at which someone does something.
The test for each metric is simple: if it moved sharply next month, would you know what to do? If not, drop it. The measures below pass that test for most restaurants and pubs, and most can be pulled from systems you already pay for.
The core measures
| Measure | How to calculate it | Where the data is | Review |
|---|---|---|---|
| Repeat-guest share | Covers from guests with a previous visit in the last 12 months, divided by all identified covers | Booking system guest profiles | Monthly |
| Review trend | Rolling 90-day average rating and number of reviews, per platform | Google, review sites | Weekly |
| Complaint rate | Complaints by email, phone or in person per 1,000 covers | Complaint log | Monthly |
| Ticket times on target | Share of food tickets served within your target time at peak | Kitchen display or EPOS | Weekly |
| No-show and late-cancel rate | No-shows plus late cancellations divided by booked covers | Booking system | Weekly |
| Dessert and coffee attach | Share of tables ordering dessert or coffee | EPOS item reports | Monthly |
| Food hygiene rating | Your current rating and next inspection due | Local authority, FSA site | After each inspection |
| Front-of-house turnover | Leavers in the last 12 months divided by average headcount | Payroll | Quarterly |
Repeat-guest share is the most direct measure of whether people come back, but it depends on capturing guest identity through bookings. If many of your guests walk in, use the share of bookings from returning guests as a proxy and say so on the scorecard. Dessert and coffee attach is a quiet signal: guests who are enjoying themselves stay for another course; guests who are not ask for the bill.
Hygiene ratings are a guest metric
Managers often treat the food hygiene rating as compliance only. Guests treat it as a reason to choose. In the Food Standards Agency's 2025 survey, 91% of consumers in England, Wales and Northern Ireland had heard of the Food Hygiene Rating Scheme and 55% had checked a business's rating in the previous 12 months[1]. Among those who knew the scheme, 52% said 4 was the lowest rating they would generally accept, and 28% had decided against a business because it did not display its sticker[1].
Put the rating and the next likely inspection on the scorecard. A drop from 5 to 3 will cost you bookings long before it shows up in reviews, and display is a legal requirement in Wales and Northern Ireland.
Reading reviews properly
- Track the 90-day rolling average, not the latest review. One bad night should not drive a menu change.
- Track volume as well as score. A falling number of reviews often means fewer covers or less engaged guests.
- Tag every review by theme: food, service, speed, value, atmosphere, booking, cleanliness. Count themes each month.
- Track your reply rate and reply time, and keep them high; guests read replies.
- Be sceptical of sudden clusters. Government research estimated that 11% to 15% of reviews in three common product categories on UK e-commerce platforms were fake[2]. Report suspected fakes through the platform rather than arguing in public.
Reviews also feed local search. Google says more reviews and positive ratings can help a business's local ranking[5]. That is a reason to ask happy guests for reviews, never a reason to buy or filter them.
Build the scorecard in an afternoon
Pick your measures
Start with five from the table. Add others only when you have a reason and an owner.
Set the baseline
Pull the last three months for each measure. That is your starting line; avoid comparing with other venues' published figures, which are measured differently.
Set a trigger for each
For example: complaint rate above your three-month average for two months running, or ticket times on target below 85% for a week. Write down who acts and how.
Put it on one page
Current figure, previous period, three-month trend, trigger and owner. If it does not fit on one page, you have too many measures.
Review it at the same meeting every month
Agree one or two actions only. Check last month's actions first.
Capturing the data without extra work
The scorecard only survives if the numbers arrive without someone spending a morning in spreadsheets. Most of the work is set-up, done once.
- Booking system: make sure every booking captures a name and either an email or a mobile number, so returning guests are recognised. Merge duplicate profiles monthly.
- EPOS: save a monthly report of covers, item counts for desserts and coffees, and any voids or refunds with a reason code. Reason codes turn refunds into a quality signal.
- Kitchen display or printer: export ticket times by hour if the system allows it; if not, have the pass write the time on a sample of tickets each Friday and Saturday.
- Complaint log: one shared sheet or form where every complaint is entered the day it arrives, with category, cause and what was offered.
- Reviews: a monthly export or a ten-minute manual count of rating, volume and themes per platform.
- Payroll: leavers and headcount each quarter.
Once set up, the monthly pull should take under an hour. If it takes longer, simplify the measures rather than skipping the review.
Give every number an owner
Measures without owners drift. Assign each one to the person who can change it: ticket times to the head chef, complaint rate and reviews to the general manager, no-shows to whoever runs reservations, front-of-house turnover to the GM with the owner. The owner explains the movement at the monthly meeting and proposes the action. Nobody is blamed for a bad month; they are expected to know why it happened and what they will try next. Over time, that habit is worth more than any individual metric, because the team learns to connect what guests say with what happens on the floor and in the kitchen.
Example: a monthly scorecard
Example (fictional): A 110-cover neighbourhood restaurant reviews its October scorecard. Repeat-guest share is steady at 38%. The review average dipped from 4.5 to 4.3, and the theme tags show 'slow' rising on Friday nights. Ticket times on target fell from 88% to 79% on Fridays only. Dessert attach dropped on the same nights. The GM's one action: cap Friday bookings at 7.30pm and add a second person on the pass. In November, Friday ticket times recover and the 'slow' tag falls away.
The example's numbers are invented. The point is the chain: a review theme, a matching operational measure and a matching spend signal, all pointing at one night and one cause. That is why the scorecard mixes guest, kitchen and sales data on one page.
Read your numbers against the market
A flat month can be good or bad depending on what is happening around you. Britain's leading managed hospitality groups grew like-for-like sales by 2.7% in September 2026, their best growth of the year[3]. Yet among trading UK accommodation and food businesses, 30.7% said domestic demand fell in September compared with August, while only 4.4% said it rose[4]. If your covers held steady while many neighbours saw demand fall, your guest experience may be doing its job.
30.7%
of trading accommodation and food businesses said domestic demand fell in September 2026[4]
Use market data for context only. Your own trend, over months, is the measure that matters.
Measures to avoid or use with care
- Social media follower counts: they rarely predict return visits.
- Average spend per head on its own: it rises when prices rise, whatever guests think.
- Survey scores with tiny response rates: a dozen replies from the happiest guests tell you little.
- Mystery-guest scores used as targets: helpful for coaching, poor as the main measure.
- Any measure nobody owns.
Complaints feed the scorecard directly; see answering complaints by email. For the bigger picture behind your covers, see eating-out habits: what the latest data shows and what diners mean by value for money.
Frequently asked questions
What is a good repeat-guest rate for a restaurant?
It varies widely by location and style; a tourist venue and a neighbourhood bistro will never compare. Track your own figure monthly and judge it by the trend rather than a published benchmark.
How many guest experience metrics should we track?
Six to eight, each with a source, an owner and a trigger for action. More than that and nobody acts on them.
Are online review scores reliable?
As a trend and a source of themes, yes. As a precise score, no: volumes are small, some reviews are fake and platforms differ. Track the rolling average and the themes.
Should the food hygiene rating be on a guest scorecard?
Yes. Many guests check it before visiting and some will not book below 4, so a lower rating affects bookings as well as compliance.
References
Food Standards Agency (2025). Food Hygiene Rating Scheme (FHRS) Food and You 2 report: wave 11 (opens in a new tab) ↑
Department for Business and Trade (2023). Fake online reviews research: executive summary (opens in a new tab) ↑
NIQ (CGA by NIQ) with RSM (2026). Bright September gives hospitality groups their best growth of 2026 (NIQ RSM Hospitality Business Tracker) (opens in a new tab) ↑
Office for National Statistics (2026). Business insights and impact on the UK economy, wave 166 dataset (opens in a new tab) ↑
Google Business Profile Help (2026). Tips to improve your local ranking on Google (opens in a new tab) ↑