Trust & Transparency
How we calculate our CSAT score
A satisfaction score you can't check is just a number on a page. Here is the formula, the responses it runs over, and what it does and doesn't tell you.
Pulled from CrewHu, the third-party survey platform our helpdesk sends through. These four totals are the entire input to the calculation below.
The formula
Every closed ticket triggers a one-click survey with three options. CrewHu labels them Great, OK, and Needs improvement. We count them like this:
Unanswered surveys are excluded entirely — they aren't counted as anything. Everything else lands in one of the three buckets, and the score covers every response on record rather than a chosen date range.
Neutral responses count against us
This is the detail worth checking on any provider's score. A widely-used alternative drops neutral responses from the calculation and reports positives ÷ (positives + negatives). Same surveys, different denominator, higher number.
Neutrals included in the denominator. This is the number we publish.
The same 12,618 responses, scored the other way. 1.81 points higher.
We publish the lower one. If you're comparing providers, it is a fair question to ask which of these two numbers you're being shown — the gap is entirely a choice about arithmetic, not about service.
What this number does not tell you
A client who closes fifty tickets contributes fifty responses. The score measures support interactions, not how many clients are happy.
Survey response is voluntary, so the sample is self-selected. That is true of every CSAT score anyone publishes, ours included.
A fast, friendly interaction can still sit inside a project that ran long. Satisfaction and results are different measurements.
Why there are no survey comments on this page
Survey responses are feedback about specific engagements, often naming a person or a system. Publishing them — even favourable ones — means publishing something a client wrote in a private channel. Aggregate counts prove the arithmetic without doing that, so counts are all this page shows.
What to ask any provider about their score
Whether or not you ever work with us, these four questions separate a measured score from a marketed one:
- Are neutral responses counted in the denominator, or dropped?
- What period does it cover — all-time, or a chosen window?
- How many responses is it calculated from?
- Is it measured per ticket, per project, or per client?
Our answers, in order: counted, all-time, 12,618, per ticket.
See what we'd find in your environment →Figures on this page are read from CrewHu when the site is built and refresh on each deploy. Method last reviewed August 2026.