Which Outbound Metrics Belong on a Weekly Dashboard (and Which Are Just Vanity)
TL;DR: A weekly outbound dashboard should answer four questions: did eligible records enter the motion, did messages reach people safely, did buyers engage meaningfully, and did that engagement create or advance pipeline. Activity counts are diagnostic inputs, not outcomes.
What outbound metrics should I track on a weekly dashboard?
Track one metric for each control point in the system: audience eligibility, delivery, engagement, conversion, and pipeline. Segment every result by play, audience, owner, channel, and launch cohort. A global reply rate can look healthy while one new sequence is burning a narrow market.
| Layer | Primary metric | Formula | Decision |
|---|---|---|---|
| Audience | Eligible enrollment rate | Eligible records enrolled / eligible records available | Find routing or capacity gaps |
| Delivery | Delivery failure rate | Failed deliveries / send attempts | Pause infrastructure or data sources |
| Engagement | Positive reply rate | Positive replies / delivered first contacts | Evaluate message and audience fit |
| Conversion | Qualified meeting rate | Qualified meetings / enrolled accounts | Evaluate whether replies become real conversations |
| Pipeline | Pipeline creation rate | New qualified opportunities / enrolled accounts | Decide whether the motion earns more coverage |
| Efficiency | Cost per qualified opportunity | Incremental program cost / new qualified opportunities | Compare the motion with other uses of budget |
Separate leading indicators from lagging outcomes
Leading indicators tell operators where a workflow is breaking before revenue data arrives. Lagging outcomes show whether the workflow produced commercial value. Keep both, but do not substitute one for the other.
| Metric | Type | Useful for | Common misuse |
|---|---|---|---|
| Eligible accounts identified | Leading | Checking audience supply | Celebrated as pipeline |
| Verified contact coverage | Leading | Diagnosing data quality | Compared across providers without matching the same sample |
| Delivery failure rate | Leading | Protecting sender health | Hidden inside a blended send count |
| Positive replies | Leading | Testing message-market resonance | Combined with referrals, objections, and opt-outs |
| Qualified meetings | Intermediate | Checking rep handoff quality | Counted before qualification |
| Qualified opportunities | Lagging | Assessing pipeline creation | Credited without a documented attribution rule |
| Closed-won revenue | Lagging | Assessing realized value | Reviewed weekly before cohorts mature |
Keep these activity metrics, but do not headline them
- Emails sent, useful for capacity and anomaly detection
- Calls completed, useful for execution coverage
- LinkedIn tasks completed, useful for workflow adherence
- Accounts researched, useful for workload forecasting
- Sequences launched, useful for change tracking
These counts become vanity metrics when they are presented without a denominator, a quality gate, or a downstream outcome. “Calls completed” is operationally useful. “More calls completed” is not a success claim unless connect quality, meeting quality, or opportunity creation improved.
Define metric contracts before building the dashboard
| Field | Question to resolve |
|---|---|
| Entity | Is the denominator a person, account, sequence enrollment, or opportunity? |
| Event | Which system event makes the metric count? |
| Window | Which dates define the cohort and the outcome period? |
| Exclusions | Are tests, internal records, duplicates, and existing customers removed? |
| Ownership | Which team owns classification and correction? |
| Attribution | What event connects outreach to the opportunity? |
| Freshness | How late can source systems arrive before the number is marked incomplete? |
Write the contract in the dashboard description. Without it, two teams can use the same label for different populations. This is especially dangerous for reply rate: the denominator may be sent messages, delivered messages, contacted people, or enrolled accounts.
Run the weekly review in 30 minutes
- Five minutes: inspect data freshness, classification backlog, and instrumentation failures
- Ten minutes: review delivery and positive replies by launch cohort
- Ten minutes: inspect qualified meetings and opportunities with owner notes
- Five minutes: choose one action for each play, continue, diagnose, change, pause, or retire
Use explicit kill, fix, and scale decisions
A dashboard should end in a decision log, not a screenshot. “Kill” means the motion is structurally invalid, for example the audience cannot be identified or the offer does not fit. “Fix” means a controllable component failed, such as data quality, copy, routing, or follow-up. “Scale” means the cohort produced enough qualified outcomes, within its observation window, to justify more volume. Set thresholds from your baseline and economics rather than borrowing universal benchmarks.
How Unify supports this operating model
Unify’s current Analytics page describes out-of-the-box dashboards for team activity, deliverability, and pipeline attribution, plus data export methods for warehouse, CRM, and other analysis workflows. The page states that performance can be viewed across sequences, plays, and reps. Use those capabilities to preserve cohort and workflow context rather than exporting a single blended activity total.
Frequently asked questions
Is open rate a useful weekly metric?
It can be a deliverability or subject-line diagnostic, but it is not a business outcome and tracking can be incomplete. Keep it below delivery, positive reply, and qualified conversion metrics.
What is the best denominator for reply rate?
Use delivered first contacts when evaluating the first-touch message, and enrolled accounts when evaluating the full account motion. State the denominator on the dashboard.
How long should a cohort stay open?
Long enough for every scheduled touch and the expected opportunity-creation window to occur. Mark immature cohorts rather than comparing them with complete cohorts.
What should RevOps do when the dashboard disagrees with the CRM?
Stop the decision, identify the event definition and freshness window, then reconcile the records. A disputed metric should not drive a scale decision.
