Most SDR dashboards are activity dashboards: dials, emails, touches, tasks completed. They answer "did people work today," which is rarely the question, and they cannot answer "what should we change," which always is.
A useful metric set does one thing: it locates the leak. Six do that.
The six, and what each diagnoses
1. Connect rate. Conversations reached divided by attempts. Diagnoses: data quality, timing, and channel fit. A low connect rate is almost never a rep problem. It is wrong numbers, wrong times, or wrong people.
2. Connect-to-conversation. Of the people reached, how many stayed past the opener. Diagnoses: the first ten seconds. This is the most coachable number in the function and the one most teams do not track at all, because it requires distinguishing a pickup from a conversation. See the first ten seconds.
3. Conversation-to-meeting. Of real conversations, how many produced a booked next step. Diagnoses: qualification and relevance. If 1 is fine and 2 is fine and this is bad, you are reaching the right people and saying the wrong thing, or reaching people with no reason to care, which is a list problem.
4. Meeting-held rate. Booked meetings that actually happened. Diagnoses: meeting quality and the confirmation process. Routinely 20 to 30% below booked. A team measured on booked rather than held will book meetings that do not happen, reliably and without anyone deciding to.
5. Meeting-to-opportunity. Held meetings that became real pipeline. Diagnoses: whether the meetings are worth having. This is the number that catches a team gaming its own targets, and it is the seam where sales development and sales have to agree on a definition.
6. Cost per opportunity. Fully loaded cost divided by opportunities sourced. Diagnoses: whether the whole function is economically viable, which is the only question the CFO is asking.
Why definitions come before targets
Every one of these has at least two common definitions, and the differences are large enough to change decisions.
| Metric | Definition A | Definition B | Typical gap |
|---|---|---|---|
| Connect rate | Any pickup | Conversation past the opener | 2 to 3x |
| Reply rate | Any response | Positive only | 2 to 4x |
| Meetings | Booked | Held | 20 to 30% |
| Per rep | Per ramped rep | Per head including ramping | 15 to 40% |
A team comparing its held-meeting number against an industry booked-meeting benchmark concludes it is failing when it is measuring something else. That is the single most common misuse of measurement in this field, and it is why borrowed benchmarks should be treated as hypotheses rather than targets.
Write your definitions down. A one-page document naming what each metric counts, and what it excludes, is worth more than any dashboard built on top of undefined numbers.
The diagnostic sequence
Read them in order. The first one that is bad is the one to work on, because each depends on the one before.
- Connect rate bad to data, timing, channel. Do not coach the opener.
- Connect-to-conversation bad to the opener. Coach this one.
- Conversation-to-meeting bad to relevance and qualification. Often the list.
- Held rate bad to the confirmation process and the quality of the ask.
- Meeting-to-opportunity bad to the definition of a qualified meeting, and
whether both sides agree on it.
- All fine, cost per opportunity still bad to the model works and the economics
do not. That is a segment or pricing question, not an execution one.
Working on step 4 while step 1 is broken is the most common wasted quarter in sales development.
What to stop reporting as a headline
- Dials and emails sent. Inputs. Report them to spot anomalies, never as
performance.
- Activity per rep leaderboards. They select for the fastest dial, which
means calling the numbers most likely to reach voicemail.
- Meetings booked without held alongside it. See metric 4.
None of these are useless. They are diagnostic detail, and treating detail as a headline is how a team ends up busy and flat.
Where this sits
This is Measurement in the Tenbound Pipeline Architecture Standard, and it is the pillar the other five depend on: you cannot tell whether a Market, Signal, Message, or Motion change worked without it. The maturity model treats consistent measurement as the gate between Assisted and Orchestrated for exactly this reason.
Related: the glossary for the definitions themselves, and SDR coaching for what to do once you know which number is the problem.