Measurement19 September 2026
What is a good sentiment score for a brand?
By Anil Salvi
Short answer
There is no credible cross-industry benchmark for brand sentiment, and any single number offered as one is invented. Sentiment scores are not comparable between tools because each vendor uses a different model, a different scale and different rules about what counts as a mention of you. They are not comparable between industries either: a bank, an airline and a snack brand sit at structurally different baselines because people post about them for different reasons. The only defensible comparison is your own score, on the same channels, against your own previous period.
- No published benchmark exists that survives scrutiny; treat any quoted figure as marketing.
- Scores are not portable between tools — switching vendors resets your history.
- Categories where people post mainly to complain sit permanently lower, and that is not a brand problem.
- Direction and cause beat level. A falling score with a named reason is more useful than a high one with none.
Why can't sentiment scores be compared across brands?
- Different models. Each vendor scores with its own classifier, and they disagree on the same post often enough to change a weekly figure.
- Different scales. Some report net sentiment (positive minus negative), some a 0–100 index, some a simple three-way split. The numbers are not interchangeable.
- Different inclusion rules. Whether a post that mentions you in passing counts at all varies by tool.
- Different channel mixes. A brand measured across review sites will look worse than the same brand measured on Instagram, because reviews skew to problems.
What should you compare your sentiment against?
- 1Your own rolling baseline, ideally 90 days, calculated per channel rather than blended.
- 2The same period last year, if you have it, to separate a real move from a seasonal one.
- 3Your named competitors, measured by the same tool on the same channels — the one genuinely fair cross-brand comparison available.
- 4The period either side of a campaign or an incident, where you are testing a specific hypothesis rather than watching a line.
How much of a move actually matters?
Less than most dashboards imply. A sentiment figure built from a few hundred mentions moves several points on ordinary variation, so small weekly wobbles are usually noise. Two things make a move worth acting on: it is large relative to your own historical variation, and it has an identifiable cause you can point at.
We have not published a threshold here — something like 'a five point move is significant' — because that depends on your mention volume, and stating one without your data behind it would be exactly the kind of invented benchmark this page warns about.
Where BrandRobin fits
Because no cross-tool benchmark exists, BrandRobin compares you against the only baseline that means anything: your own 90 days, per channel. Low-confidence calls are flagged rather than counted quietly, and your corrections are kept.
- A 90-day rolling baseline per channel, not a fixed threshold
- Sample size shown next to every figure, so a small sample cannot pose as a trend
- Correct a mis-scored mention and the correction sticks for similar ones
- Sentiment on every plan from $49 a month; themes and campaign windows from $129
Frequently asked questions
- Is 70% positive a good score?
- It is not answerable without knowing the tool, the channels and the category. A 70% positive figure drawn mostly from Instagram and the same figure drawn mostly from Trustpilot describe very different situations. Compare it against your own last quarter instead.
- Why did our sentiment drop when nothing bad happened?
- Most often the channel mix changed — a review site started contributing more mentions, or a campaign brought in a wave of neutral posts that the model scored conservatively. Check volume by channel before assuming a reputation problem.
- Should sentiment be a KPI?
- As a tracked measure with a named owner, yes. As a target with a number attached, rarely: it is heavily influenced by things outside marketing's control, and teams optimising it tend to start excluding inconvenient channels.
- How accurate is automated sentiment scoring?
- Good on plain statements, weak on sarcasm, in-jokes and industry shorthand — true of every vendor including the expensive ones. What varies is whether a tool flags its low-confidence calls or quietly counts them, which matters more than the headline accuracy claim.
How this was put together
This page deliberately publishes no benchmark figure. We looked for a credible cross-industry sentiment norm and did not find one we would stand behind: the numbers in circulation come from individual vendors measuring their own customer bases with their own models, which makes them incomparable by construction. Saying so is more useful than picking one.
Go deeper on the blog
What people ask next
What is share of voice and how do you calculate it?
Your share of the public conversation about a named set of brands. Simple arithmetic; the judgement is in what goes into the denominator.
Read the answerIs share of voice the same as market share?
No. One counts conversation, the other counts revenue, and they routinely move in opposite directions.
Read the answerWhat is social listening?
Reading what people say in public about your brand, your competitors and your category, then acting on it — as opposed to counting how often your name appears.
Read the answer