Most RevOps involvement in outbound starts after the problem. Reply rates have halved, nobody can say why, and someone asks for a dashboard.
The dashboard will not answer it, because the data needed to answer it was never captured. That is the actual job: making sure that when the question arrives, the answer exists.
There are four questions. Everything below exists to keep them answerable.
- Which outbound produced which revenue?
- What does a meeting cost us, by channel?
- When a metric moves, what moved it?
- Is the system supplied, or is it running out of leads?
Question one: the attribution chain
The chain has four links and it breaks at the third.
Lead source → campaign → reply → opportunity.
Links one and two usually survive, because the outreach platform holds them. Link four survives, because the CRM holds it. Link three is where it fails: the reply happens in one system, the opportunity is created in another, and the person creating it does not carry the campaign across.
Six weeks later you have 40 opportunities from outbound and no way to say which campaign produced any of them.
The seven fields that must survive every handoff:
| Field | Why it matters | Where it dies |
|---|---|---|
| Lead source | Database search vs signal vs imported list | Overwritten by the CRM’s own source logic |
| Campaign ID | The only link to the copy and the list | Never written on manual opportunity creation |
| Channel of first touch | Attribution by channel | Collapsed into “Outbound” |
| Channel of reply | Which channel actually worked | Rarely captured at all |
| Sequence step of reply | Tells you how many steps you need | Not exposed by most tools’ CRM sync |
| Signal type, if any | The highest-value segmentation you have | Not a CRM concept, so not mapped |
| Date of first touch | Cycle length from first contact | Replaced by “created date” |
Two of those deserve emphasis.
Channel of reply, separate from channel of first touch. A sequence that opens on LinkedIn and gets the reply by email is a very different machine from one that does both on LinkedIn. Without the distinction you will attribute the result to the wrong channel and then cut the wrong one.
Sequence step of reply. This is the field nobody captures and the one that most changes behaviour. If 80% of replies come from steps one and two, your six-step sequence is costing sender reputation for nothing. You cannot know that without the field.
Make it non-optional. Not a required field on a form someone can bypass, but written by the integration at the moment the opportunity is created. If a rep can create an outbound opportunity by hand with those fields blank, within a quarter most of them will be blank. Automate the write, seven automations worth building includes the one that does this.
Question two: cost per meeting, properly
Most teams compute software spend divided by meetings. That number is wrong by a factor of three or more, and it is wrong in a way that favours the most expensive channel.
Everything that belongs in the model:
- Platform subscription, apportioned by channel
- Mailbox and domain costs, and dedicated IP where applicable
- Data credits consumed, including the ones that returned nothing
- LinkedIn seats or rented profile costs
- Telephony: numbers plus per-minute
- Human time, costed at loaded rate, this is usually the largest line and the one always omitted
- A share of the setup cost, amortised
Human time is the one that changes conclusions. Cold calling with a $0.02-per-minute rate looks like the cheapest channel until you price the 45 minutes of dialling per connected conversation. Email looks expensive on infrastructure until you notice it needs almost no time per send.
Build it per channel per month and let it be ugly. An approximate model that includes rep time beats a precise model that excludes it. What an outbound stack costs has the line items; what a B2B lead costs has the data side.
Then track the ratio over time rather than the absolute. The absolute is for budget conversations. The trend is what tells you something broke.
Question three: definitions, before dashboards
Four definitions that every team believes are shared and almost none are. Write them down, because without them your trend lines measure reporting behaviour rather than performance.
What counts as a reply. Does an out-of-office count? An auto-acknowledgement? A bounce that contains text? If your platform counts OOOs as replies, your reply rate has a seasonal pattern that means nothing.
What counts as a positive reply. The one that matters and the one most loosely defined. “Interested” versus “asked a question” versus “referred us to someone” are three different outcomes. Pick categories, make them exhaustive and mutually exclusive, and audit a sample monthly. Categorisation drifts quietly, reps grade generously in a bad month.
What counts as a meeting. Booked, held or qualified. A booked-meetings number with a 30% no-show rate is a different metric from held meetings, and teams switch between them without saying so. Track booked and held separately and report the gap, because the gap is itself a signal about targeting quality.
What counts as outbound. A prospect who was emailed in March, ignored it, then filled in a form in July. Marketing will claim it; outbound touched it first. Decide the rule in advance, a time window is the usual answer, and apply it consistently. Arguing about it after the quarter closes is how RevOps loses credibility.
Question four: supply
The metric almost nobody tracks and the one that predicts next quarter.
Days of lead supply remaining, per campaign: unsent leads divided by daily send volume.
A campaign at eleven days means someone needs to build a list this week. Nobody notices until sends drop, which is three weeks after the point where it was cheap to fix. By then there is a hole in the pipeline that will show up in revenue a quarter later.
Two related supply metrics:
List decay rate. What percentage of a list is stale after 90 days. Around 2–3% of B2B contacts change role monthly, which compounds faster than people expect. How fast lists decay covers the arithmetic.
Sending capacity utilisation. Sends as a percentage of what your mailboxes can safely carry. Running at 95% means no headroom for a campaign launch; running at 40% means you are paying for infrastructure you are not using.
The three leading indicators
Lagging metrics tell you what happened. These three tell you what is about to.
Days of lead supply. Covered above. Predicts send volume three weeks out.
Bounce rate by list source. Rising bounce on one source predicts a deliverability problem before placement moves. Segment it by source or the signal is buried in the average.
Reply-to-meeting ratio. If reply rate holds and this falls, the problem is after the reply, usually response time. This is the single most diagnostic ratio in outbound and almost nobody has it on a chart.
Where the tooling fights you
Three recurring problems, with what actually works.
The CRM was not built for this. Objects assume an inbound funnel with one touch. Outbound has forty touches across four channels against six contacts at one account. Resolve it by syncing outcomes rather than activities: do not push every open and profile view into the CRM. It produces an unreadable timeline and buries the three events that mattered. Whether you need a CRM for outbound covers the threshold.
Multiple tools, multiple truths. The outreach platform, the CRM and the dialler each report meetings and each gives a different number. Pick one system of record per metric, write it down, and stop reconciling. Reconciliation is infinite work with no output.
Metrics that cannot be segmented. An aggregate reply rate across four channels and nine campaigns is almost useless, one bad campaign hides inside a good average. Insist on segmentability by channel, campaign, list source, signal type and sequence step before you build any reporting at all.
The quarterly audit
Thirty minutes, four checks.
- Sample 20 outbound opportunities. How many carry all seven fields? Below 90% means the automation is broken or being bypassed.
- Sample 20 replies graded positive. Do you agree with the grading? Disagreement over 20% means the categories need rewriting.
- Recompute cost per meeting with current prices. Tool prices and headcount both change.
- Check supply across every active campaign. Anything under three weeks gets flagged now.
If you do nothing else from this post, do the first two. Field integrity and definition integrity are what make every other number mean something, and both decay silently.
The short version
Your job is not the dashboard. It is making sure four questions stay answerable: which outbound produced which revenue, what a meeting costs by channel, what moved when a metric moved, and whether the system is still supplied.
Seven fields have to survive every handoff, written by the integration rather than by a person. Sequence step of reply and channel of reply are the two nobody captures and the two that most change decisions.
Agree four definitions before building anything: reply, positive reply, meeting, and what counts as outbound.
Track days of lead supply, bounce rate by list source, and reply-to-meeting. Those three move before revenue does.
Next: whether you need a CRM for outbound, or what an outbound stack costs.
Frequently asked questions
What should RevOps measure for outbound?
Beyond reply and meeting rates: days of lead supply per campaign, bounce rate segmented by list source, reply-to-meeting ratio, and cost per meeting by channel including loaded rep time. The first and last are the ones most often missing.
Why does outbound attribution break?
Because the reply happens in the outreach platform and the opportunity is created in the CRM, and whoever creates it rarely carries the campaign, channel and source fields across. Automating that write at creation time is the fix.
How do you calculate cost per meeting for outbound?
Include platform cost apportioned by channel, mailboxes, domains and IPs, data credits including failed lookups, LinkedIn seats, telephony, amortised setup, and rep time at loaded rate. Rep time is usually the largest line and the one most often left out.
What is the most useful outbound metric nobody tracks?
Days of lead supply remaining per campaign, unsent leads divided by daily send volume. It predicts a pipeline gap three weeks before send volume visibly drops.