A practical weekly reporting template with matched cohorts, clear denominators, meeting outcomes and traceable source data.
A useful LinkedIn outreach report lets a client follow a number back to the work behind it. They should be able to see what happened, which prospects moved forward, where information is missing, and what the team will do next.
That takes more than a row of totals. Thirty accepted invitations can be a correct weekly count and an incorrect numerator for that week's acceptance rate. A meeting on the calendar can still become a cancellation. A profile's connection count describes its network, without identifying the campaign that created those connections.
This guide gives agencies and outreach teams a reporting framework, a worked example, and a weekly template. Product details were checked on September 30, 2026. The definitions and workflow recommendations below are our editorial approach; every numerical example is illustrative, not an AllProfiles customer result or a performance benchmark.
Define the events before designing the dashboard
Agree on a short metric dictionary with the client. Keep it beside the report so that a new operator cannot silently change what a number means.
| Metric | Suggested counting rule | Record that supports it |
|---|---|---|
| Invitations sent | Successful invitation events during the period; track unique recipients separately | Prospect reference, sender and sent time |
| Invitations accepted | Recorded acceptances during the period | Acceptance time and original invitation, where available |
| People messaged | Unique recipients of the specified outbound message channel | Recipient, channel and first qualifying message |
| People who replied | Unique people with a human response in the defined reporting scope | Conversation reference and response time |
| Positive replies | People whose replies meet the agreed interest criteria | Classification and supporting conversation |
| Qualified conversations | Prospects meeting the client's written qualification criteria | Qualification evidence, reviewer and date |
| Meetings booked | Distinct meetings first booked during the period | Stable meeting reference and original booking time |
| Meetings held | Meetings confirmed completed during the period | Meeting outcome and actual meeting date |
Follow-up messages count as additional messages, but they should not turn one person into several people in a unique-recipient denominator. Keep automatic responses separate from human replies. Count a first positive reply once per prospect and campaign; preserve later status changes in the record.
These metrics do not always form a perfectly descending funnel. LinkedIn explicitly says someone can reply to an invitation message without accepting the connection invitation. A reply therefore does not prove a connection was established. Read LinkedIn's invitation-response guidance.
Use separate event dates and identifiers instead of forcing every conversation through an assumed sequence.
Separate weekly activity from send-cohort results
An activity report answers: what happened during these dates?
A cohort report answers: what happened to the prospects first contacted during these dates, by a stated observation cutoff?
Both views are useful. Mixing them produces percentages that look precise but describe the wrong population.
Illustrative example — invented data for explanation only:
| Event recorded during one week | Count |
|---|---|
| Invitations sent that week | 100 |
| Acceptances recorded that week | 30 |
| Acceptances from invitations sent before that week | 20 |
| Acceptances from invitations sent that week | 10 |
The activity summary correctly says 100 invitations sent and 30 accepted during the week.
But 30 divided by 100 is not the acceptance rate for this send cohort: twenty acceptances came from older invitations. For the hundred invitations sent this week, the observed result is 10 divided by 100 = 10% accepted as of the cutoff.
Even that result is provisional. Someone contacted Monday has had longer to respond than someone contacted Friday. For a fair comparison, choose a consistent observation window, such as acceptance within 14 days of each invitation, and compare cohorts only after every included invitation has had that window. Fourteen days is an example of a reporting convention, not a LinkedIn limit or recommended sending schedule.
If your source cannot link an acceptance to its original invitation, show the activity count and mark cohort conversion as unavailable. Do not manufacture the missing link from aggregate totals.
Put the denominator next to the percentage
Write each rate so that another person can reproduce it:
- Invitation acceptance rate: invitations from the stated send cohort accepted within the stated window, divided by invitations in that cohort.
- Unique-person reply rate: unique people in the specified message cohort who gave a human reply within the window, divided by unique people messaged in that cohort.
- Qualified-conversation rate: unique people from the defined outreach cohort who met the agreed criteria by the cutoff, divided by unique people in that cohort.
Keep invitation messages, direct messages and InMail distinguishable. Specify how cross-channel replies are attributed before combining them. If a meeting follows both LinkedIn outreach and an email introduction, apply the agreed attribution rule and retain both touchpoints rather than giving each channel sole credit.
For a zero denominator, show N/A. If the denominator is missing, show unknown. Neither situation means a 0% conversion rate.
Combine the underlying counts, not the displayed rates
Another illustrative example: one non-overlapping group has 10 acceptances from 20 invitations, while another has 20 from 100. Their combined acceptance rate is 30 / 120 = 25%. Averaging the displayed 50% and 20% would give 35%, overstating the combined result.
Before combining groups, confirm that their definitions and observation windows match. For a client-level unique-person metric, also remove duplicate prospects appearing under multiple operators. Keep those duplicates visible in the operational log so the team can fix overlapping assignments.
Show the difference between interest and a sales handoff
“Sounds interesting” can justify a positive-reply label. It does not establish buying authority, company fit or a relevant current need.
An illustrative qualification rule might require three things: the company fits the agreed target market, the contact describes a relevant problem or responsibility, and both sides agree on a useful next step. Your client may need different criteria. Write them down before the campaign begins and record any changes with an effective date.
CRM terminology also needs an explicit mapping. HubSpot, for example, distinguishes a sales-qualified lead from an opportunity associated with a deal, and supports customized lifecycle stages. Those are CRM definitions, not universal LinkedIn labels. See HubSpot's lifecycle-stage documentation.
A useful handoff contains the prospect reference, a short qualification note, the responsible person, and the next action with a due date. A client should be able to act on the record without rereading an entire conversation. Share only the conversation material appropriate to that client.
Track meetings through their outcome
Keep the original booking date, current scheduled date and outcome as separate fields. A meeting booked this week may happen next week; a meeting held this week may have been booked last month.
For this reporting framework, a reschedule updates the same meeting record instead of creating a second acquisition win. Record cancellations and no-shows explicitly. A calendar invitation alone does not confirm attendance.
This distinction is supported by common CRM data models: HubSpot documents separate meeting start/end times and completed, cancelled, no-show and rescheduled outcome properties. Its documentation also notes that connected scheduling integrations may use different properties. Review the meeting properties before mapping your data.
Do not divide meetings held this week by meetings booked this week and call the result a show rate: the groups can be different. For attendance reporting, choose a defined set of meetings with scheduled dates in the period, establish how reschedules and cancellations are treated, and calculate against that same set. Disclose unresolved outcomes instead of silently treating them as no-shows.
Make source coverage and freshness visible
Every report needs an as-of time, a timezone, and a coverage statement. A dashboard refreshed this morning may still contain a provider snapshot from two days ago.
LinkedIn's Sales Navigator usage report is a useful example of why labels matter. Its metrics update daily and may refresh at different times. Its “Messages sent” metric excludes messages sent from LinkedIn.com and connection requests. Usage Reporting is available to Advanced and Advanced Plus admins, subject to LinkedIn's documented access conditions. Check the official report definitions.
Read the definition of each source field before presenting it as the team's total activity. If you are choosing a subscription for the work, our Sales Navigator versus Premium Business guide explains the purchasing distinctions.
A connection snapshot is not an outreach activity log
A current network total cannot tell you how many campaign invitations were sent, who replied, or which meeting belongs to a campaign. Even the difference between two network totals does not identify the contributing events or their owners.
For example, the AllProfiles private client portal displays assigned account inventory and available connection snapshots. That view helps reconcile which profiles belong to a client and when their metadata was checked. Outreach events, conversation classifications and meeting outcomes need their own records; do not infer them from those snapshots.
When a source is missing, retain the last successful value with its timestamp or display “not available.” Avoid turning an absent daily entry into zero activity. Show corrections separately so clients can understand why a previous total changed.
A weekly client-report template
Start the report with these fields:
Client · Reporting period · Timezone · Prepared by · Data current through · Covered campaigns and operators · Known missing sources or days
Then use this structure:
| Section | What the client should see |
|---|---|
| Activity during the period | Sent invitations, recorded acceptances, unique people messaged and replying, first positive replies, newly qualified conversations, meetings booked and held |
| Cohort results | Send dates, observation window, matched outcome count, eligible denominator, rate and whether the result is still provisional |
| Qualified conversations | Prospect reference, qualification evidence, owner, next step and due date |
| Meetings | Meeting reference, first booking date, scheduled date, current outcome and follow-up owner |
| Data quality | Source and last refresh, missing records, late entries and corrections since the previous report |
| Next decision | One supported interpretation, one action, a responsible person and a review date |
Keep a supporting event ledger with a stable prospect identifier, client, campaign, operator, event type, event time, source reference and classification. For meetings, include a stable meeting identifier and outcome. Store the time an event occurred separately from the time someone entered it; a late entry should not masquerade as today's work.
A spreadsheet can implement this structure before you automate it. Give clients a concise summary and access to relevant supporting records, without exposing other clients' data or internal credentials.
Turn the report into a decision
A useful commentary names an observed issue and a specific response. For example: “Two qualified conversations are waiting for a client answer; the account owner will resolve them by Thursday.” This illustrative statement gives the client something actionable. “Engagement is improving” does not explain what to do.
Before sending the report, check five things:
- Every percentage has a visible numerator, denominator, cohort and observation window.
- Events and unique people use separate labels; duplicates and reschedules follow one written rule.
- Booked, held and qualified outcomes have supporting records.
- Missing sources, stale snapshots and late corrections are visible.
- The client has a clear next action, owner and review date.
If you are setting up the outreach work itself, use our two-profile pilot guide for responsibilities and evaluation planning. Then use this reporting structure consistently across the operators and profiles assigned to the client.
