A note on this page. It replaces a URL that delivered a Tenbound benchmark report. That report's underlying data is not available to restate accurately, so its figures are not reproduced here. Publishing benchmark numbers that cannot be checked would be worse than publishing none. What follows is how to get benchmarks you can actually use, which is the more durable answer.
Why most published benchmarks are not comparable
Not because they are dishonest. Because the definitions vary and are rarely published.
- Connect rate. Any pickup, or a conversation past the opener? Those differ
by a factor of three in practice.
- Reply rate. Any response including negatives, or positive only? Including
out-of-office?
- Meetings. Booked, or held? The gap is routinely 20 to 30%.
- Per rep, per what? Per ramped rep, per head including ramping ones, per
full-time equivalent?
A team comparing its held-meeting count to someone else's booked-meeting count concludes it is underperforming when it is measuring something else. That is the single most common misuse of benchmarks in this field.
The definitional detail is in the sales development glossary, and it exists because these arguments are constant.
Build your own baseline
Two quarters. Nothing shorter survives seasonality.
1. Write the definitions first. Before collecting anything. One page, agreed by the SDR manager and the AE leader, covering connect, reply, meeting set, meeting held, and next step.
2. Instrument what you already have. Most teams have the data and have never extracted it in a comparable form. Do not buy anything for this.
3. Segment from the start. Inbound and outbound, by tier, and by rep tenure. A blended baseline is unusable, because it moves with mix and hides where the variance is.
4. Record the conditions. What changed in the period: territory changes, product launches, a deliverability incident, a quarter with three public holidays. A baseline without an event log cannot be interpreted later.
5. Publish the distribution, not the average. Quartiles. An average connect rate across a team where two reps carry the phone describes nobody.
Which external numbers are safe to borrow
Roughly in order of reliability:
- Your own prior periods. Always best.
- Aggregated operational data, where a provider measured what happened
rather than asking people to recall it.
- Published research with a methodology section and a stated definition.
- Vendor-sponsored surveys, with the four checks in
- Conference-stage numbers. Anecdote unless sourced.
Use benchmarks for direction, not for grading
The legitimate use is orientation: is our meeting-held rate in the same territory as others, or an order of magnitude away? An order-of-magnitude gap is worth investigating whatever the definitional noise.
The illegitimate use is a target. A rep graded against an external number measured differently is being graded on someone else's definition, which is both unfair and uninformative.
Where this sits
Benchmarking is Measurement in the Tenbound Pipeline Architecture Standard, and it is where the definitional discipline in the rest of the framework pays off. A team with written stage definitions can benchmark itself against itself, which is the only comparison guaranteed to be valid.