Reputation Score

What is Reputation Score?

A reputation score is a single number that summarizes how your business is perceived online. It combines several signals, typically your star ratings, how many reviews you receive, how recent they are, how consistently you respond and what guests say, into one index.

Because the score is a composite, its meaning depends entirely on how it is built. Reputation management vendors each define their own version, and many restaurant groups create an internal one. There is no industry standard formula, which makes it essential to know what goes into the number you're looking at.

For a multi-location restaurant brand, a reputation score is useful because it turns a large amount of data into one comparable figure per location. It sits within the wider practice of online reputation management and builds on the star rating each platform calculates.

How a reputation score works

Most reputation scores follow the same three steps, whatever the vendor:

1. Choosing the components. Each component measures one aspect of reputation, such as average rating or response rate.
2. Normalizing each component. Each input is converted to a common scale, usually 0 to 100, so a star rating and a percentage can be combined.
3. Weighting and adding. Each normalized score is multiplied by its weight, and the results are added together.

Reputation score = Σ (component score × weight) ÷ Σ weights

Here is a worked example for one location, using illustrative weights: rating 35%, review volume 15%, recency 15%, response rate 15% and sentiment 20%. These weights are an example, and your own choice may differ.

The points add up to 72.525, so this location's reputation score is 73 out of 100.

The table shows the components as well as the total, which is what makes the score useful. Volume is the weakest component here, and it carries the largest available gain: 50 points of headroom at a 15% weight is worth up to 7.5 points.

Reputation score calculator

Enter one location's figures and, if you like, change the weights. The weights below are an illustration of one reasonable starting point, and the calculator rescales them if they don't add up to 100.

The calculator caps volume and recency at 100 once you reach your target, and it converts the star rating so that 1 star scores 0 and 5 stars score 100. Different normalizations are equally valid, which is one reason vendor scores differ.

Reputation score components

Star rating component

Your average rating is usually the heaviest input, because it is what guests see first. Each platform calculates it differently: Google's help page on review scores describes a plain average, and Uber's ratings policy gives priority to the most recent merchant ratings.

When you combine platforms, weighting each by its share of reviews or by its importance to your business keeps one small platform from dominating the score.

Volume and recency components

Volume shows how much evidence stands behind your rating, and recency shows whether it reflects the current experience. A location with a 4.6 rating and no reviews in three months is in a weaker position than the rating alone suggests.

Small samples deserve particular care. Evan Miller's widely cited essay How Not To Sort By Average Rating (2009) explains why a raw average from a handful of ratings is unreliable, which is why some scores discount ratings backed by few reviews.

Response component

Response rate and response time are the components most directly under your team's control.

EN Group, for example, achieved a 98% customer response rate using Momos.

Sentiment component

Sentiment measures what guests actually write, which a star average can't capture. Sentiment analysis classifies each review or each topic within it, so the score can reflect, for example, rising complaints about wait times before they lower your rating.

Types of reputation score

Vendor reputation scores

Reputation management platforms calculate their own scores from the data they collect. These are convenient and often come with benchmarks, though the exact formula is usually unpublished.

Because each vendor's components and scales differ, a vendor score is best compared only with other scores from the same vendor.

Internal reputation scorecards

Many restaurant groups build their own score from the metrics they already track. The advantage is transparency: everyone knows exactly what moves the number. Dashboards that let you customize your headline metrics make it practical to show an internal score next to its components.

Location-level and brand-level scores

A location-level score shows each manager where to focus. A brand-level score, usually a weighted average of locations, gives leadership one trend line. Both are useful, and the location view tends to drive more change.

Reputation score vs. star rating

Your star rating is the number guests act on, so it deserves attention in its own right. The reputation score adds context, showing whether that rating is well supported, current and actively managed, and it pairs naturally with private measures such as CSAT and NPS.

How to build a reputation score for your brand

A useful internal reputation score is simple enough for every general manager to understand. These steps reflect common practice:

1. Deciding what the score is for. A score for comparing locations needs inputs every location can influence, such as response rate and review volume.
2. Choosing four to six components. More components make the score harder to explain and each one less influential.
3. Normalizing consistently. Setting targets that are realistic for your locations, and documenting how each input becomes a 0–100 score, keeps results stable.
4. Choosing and documenting the weights. Weights reflect your priorities. Testing a few alternatives shows how much the ranking of locations depends on them.
5. Keeping the formula stable. Changing weights every quarter makes trends meaningless, so changes are best made rarely and recorded.
6. Always showing components next to the total. An operator report that lists each location's score with its components tells managers what to work on.

How to interpret a reputation score

A reputation score is most informative as a comparison: between locations, and over time for the same location. A five-point gap between two locations matters more when you can see which component explains it.

Month-to-month movements at low-volume locations can come from a handful of reviews, so a rolling three-month view smooths out much of that noise. Pairing the score with sales data shows whether changes in reputation come with changes in visits.

Published research supports paying attention to the rating component in particular. Michael Luca's Harvard Business School working paper Reviews, Reputation, and Revenue: The Case of Yelp.com (2011) found that a one-star increase in Yelp rating led to a 5 to 9% increase in revenue for independent restaurants.

Limitations of a reputation score

• Arbitrary weights. The weights are judgments, and a different set can reorder your locations.
• Masking. A strong response rate can offset a falling star rating in the total, hiding the problem guests actually see.
• Opaque vendor formulas. An unpublished method makes it hard to know what to improve or why the score moved.
• Platform rules underneath. Inputs inherit each platform's filtering; Yelp, for example, leaves reviews that aren't currently recommended out of its star rating.
• Gaming. Any score tied to bonuses invites shortcuts, such as fake or incentivized reviews, which the FTC's 2024 rule on fake reviews prohibits.

How to improve your reputation score

Improving a reputation score means improving its components, starting with the one that offers the largest gain at each location.

1. Raising response rate and speed. Replying to every review, with help from an AI review manager that drafts replies in your brand's voice, lifts the response component quickly. Lee's Famous Recipe Chicken boosted review response rates by 24%.
2. Growing volume and recency fairly. Inviting every guest through review solicitation and review boosting adds fresh reviews. Caribou Coffee achieved 61% more reviews across 57 locations.
3. Fixing what lowers ratings and sentiment. Recurring themes in negative reviews point to operational fixes, and a thoughtful negative review response with customer recovery can win guests back.
4. Keeping listings accurate. Correct hours and menus prevent avoidable complaints from guests who arrive to find you closed.

Best reputation score tools for restaurants

When choosing a tool to calculate or track a reputation score, the most important factors are coverage of every platform your guests use, transparency about how any score is calculated, and reporting that breaks the score down by location and component.

It helps to ask each vendor how its score is built and whether you can see the components.

Common reputation score mistakes

• Treating a vendor score as a universal standard. Scores from different formulas can't be compared.
• Reporting the total without components. Managers can't act on a number they can't break down.
• Changing the weights often. Frequent changes break the trend line.
• Over-weighting volume. Rewarding volume heavily can push teams toward review requests that break platform rules.
• Ignoring low-volume locations' noise. A few reviews can swing a small location's score.
• Forgetting the guest's view. Guests never see your reputation score; they see your star rating and your replies.

History of the reputation score

Combining several indicators into one index is a long-established technique. The OECD and the European Commission's Joint Research Centre set out the method in their Handbook on Constructing Composite Indicators (2008), including steps for normalizing indicators, weighting them and testing how sensitive the result is to those choices.

Online, the idea took hold as reviews spread across Google, Yelp, Tripadvisor and delivery apps. A brand with dozens of locations and several platforms needed one way to see where it stood, and reputation management software began offering a single score.

Today, several vendors publish proprietary scores. Reputation, for example, publishes a Reputation Score built from components such as review sentiment, volume, recency and response. Other platforms use different components, weights and scales, and none of these formulas has become a shared standard.