Principles
Queue Bench benchmarks are based entirely on publicly observable data. We do not accept vendor-provided data, run internal tests, or estimate figures that are not published.
Data collection
For each benchmark run:
- We navigate to the vendor’s public pricing or features page
- We record the specific figure, the URL, and the date
- We store a direct quote from the page
- We flag any figure that requires contacting sales as “Not published”
Ranking rules
- Lower is better metrics (AI price per outcome, seat cost): vendors with lower published values rank higher.
- Higher is better metrics (channel coverage): vendors with higher published values rank higher.
- Vendors with “Not published” status always rank below every published value, regardless of direction.
- Ties are shown as “2=” with no invented tie-break.
- Units are never mixed across a single leaderboard.
The three launch metrics
B1: AI Price per Outcome. The published cost per AI-resolved ticket, charged by the vendor. Only per-outcome or per-resolution pricing qualifies. Seat-cost estimates are not used.
B2: Seat Cost. The cheapest published plan that includes both email and live chat, priced per seat per month on annual billing.
B3: Channel Coverage. Count of the 12 tracked channels confirmed as included in the base plan. Add-ons and unconfirmed channels score 0.
Not published
“Not published” is a scoring statement, not a criticism. It means the figure was not on the public page on the run date. A vendor can change this by publishing the price.
Corrections and disputes
Email hello@queuebench.com with a URL and quote showing a different figure. Verified corrections update the run within five business days.