TL;DR
- This AI outbound campaign report template fits on one page. It shows who we reached, what happened, what it cost and what comes next.
- Lead with reached calls and outcomes, not dials.
- Agree metric definitions before launch.
Copy this one-page AI outbound campaign report
Every figure below is an illustrative placeholder with round numbers. Replace them with your own results.
1. Headline and decision asked
- Example: "The win-back campaign reached 1,500 customers and booked 240 repayment plans. We ask leadership to approve a second wave of 5,000 contacts."
2. Campaign setup
| Field | Example entry |
|---|---|
| Type | Win-back (survey, lead follow-up, cold outreach and win-back are the campaign types) |
| Goal | Book a repayment-plan call |
| Audience | 5,000 lapsed customers |
| Calling hours | Weekdays, 9:00 to 18:00 local |
| Retry rules | Up to 3 attempts, 1 day apart |
| Consent marking | Marked on every contact; do-not-call list applied |
| Voice and language | One AI voice, English and Yoruba |
| Dates | 1 to 14 March |
3. Funnel
| Stage | Example |
|---|---|
| Contacts loaded | 5,000 |
| Dialled | 4,800 |
| Answered | 1,900 |
| Reached | 1,500 |
| Real conversation | 1,200 |
| Goal outcome | 240 |
| Opt-outs | 30 |
Define "reached" once, in writing. Calls where only the AI spoke do not count as reached. In the example, 400 answered calls had no customer speech, so reached is 1,500, not 1,900. A carrier's temporary refusal does not use up a retry attempt, so retry counts stay honest.
4. Outcome breakdown (of 1,200 conversations)
- Interested: 240
- Callback requested: 180
- Wrong number: 100
- Declined: 680
5. Comparison or control group
- Voice AI group: 1,000 contacts, 6% reached the goal.
- Holdout group: 1,000 contacts, no call, 2.5% reached the goal.
- Lift: 3.5 percentage points.
6. Cost and value
- Total campaign cost: 6,000.
- Cost per goal outcome: 25 (6,000 divided by 240).
- Our cost to acquire or serve one customer: 80.
7. Quality and risk (placeholder figures)
- Transcripts reviewed: 120 (10% sample).
- Handovers to a human: 90.
- Complaints: 5.
- Opt-outs: 30.
- AI disclosed at call start: 97% of sampled calls (placeholder, set your own target).
- Silent or AI-only calls: 400.
8. Learnings and next step
- Learning: answer rate rose in the 17:00 to 18:00 hour.
- Next step: shift calling hours later and run wave two.
- Owner: Head of Collections. Date: 28 March.
Where the data comes from
Your platform should give you most of these numbers. Insight7's outbound campaigns are in beta. They have a live dashboard of calls in progress, and a reviewer can read each transcript when the call ends. Yes/no answers and outcomes are pulled from the transcript. A campaign scorecard is also produced when the campaign ends, but the detail is still in beta, so check it in the app before you rely on it.
Fill in the numbers with three formulas
Work out each rate from the funnel, so leaders can check your maths.
- Reached rate: reached divided by dialled. In the example, 1,500 divided by 4,800 is about 31%.
- Cost per goal outcome: total cost divided by goal outcomes. Here, 6,000 divided by 240 is 25. Compare it with your cost to acquire or serve a customer, 80 in the example.
- Lift: goal rate in the called group minus goal rate in the holdout. Here, 6% minus 2.5% is 3.5 points. Multiply the lift by the audience size to estimate the extra outcomes the calls caused.
Fill the page in this order: setup, funnel, control, cost. Then write the headline last. The decision you ask for should follow from the numbers beneath it.
Adapt the page for your audience and risk
Change the two lead metrics to match the campaign type. Then change the risk block to match the reader. The other blocks stay the same.
| Campaign type | Lead with |
|---|---|
| Customer survey | Completed surveys, reached rate |
| Lead follow-up | Real conversations, meetings booked |
| Cold outreach | Reached rate, opt-out rate |
| Win-back | Customers reactivated, cost per outcome |
| Payment reminders | Payments made before the due date, reached rate |
Pick the lead block by reader
- A CFO wants cost per goal outcome, shown next to your cost to acquire or serve a customer.
- A compliance or risk head wants consent marking, disclosure, opt-outs, human handovers and the share of transcripts reviewed.
- A CX head wants complaints and customer sentiment, taken from the transcripts.
Show regulated buyers the guardrails
In banks and lenders, the main blocker is fear of an AI agent that misbehaves in public. These buyers want proof from real deployments, not a better demo. Fill the risk block with your handover rules and your review sample. Buyers also ask three things: is the handover polite, what happens to a sensitive issue or a disputed amount, and does anyone chase a handover nobody picked up?
Report disclosure and consent by region
List the rules that apply to your calls:
- EU: Article 50 of the AI Act has applied since 2 August 2026. Any AI that talks to people, including automated phone systems, must say it is AI from the first interaction.
- US: AI-voice calls need consent under the TCPA, following the FCC's 2024 ruling. A proposed rule to say "this is AI" at the start was still not final as of mid-2026.
Customers also expect a route to a person. In a Gartner survey of 3,566 customers, 87% said a company using GenAI for service must offer a way to reach a human. That is why the handover count belongs on the page.
Keep it to one page
For campaigns that run all year, send a monthly version with a trend column on the funnel and the setup table cut to one line. For a pilot-end version, add an appendix of sample transcripts.
Related reading: AI Voice Calls vs SMS vs Email for Customer Reactivation Campaigns Compared by Response and Conversion, AI Voice Agent Pilot Plan Template for a Two-Week Proof of Concept and AI Voice Agent ROI Calculator With Worked Examples for Customer Service and Collections.
Questions people also ask
What should an AI outbound campaign report include?
Eight blocks: the headline and decision, setup, funnel, outcomes, control result, cost and value, quality and risk, and next step. Put the decision first. Cut any figure that does not help leadership choose.
How long should a campaign report for leadership be?
One page. Leaders read the top third and decide from it. Keep the funnel, cost and risk on the page. Move sample transcripts and call-level detail to an appendix.
What is the difference between answered calls and reached calls?
Answered means a line picked up, which can include voicemail or silence. Reached means a real person spoke. Report both, so leaders can see how many answered calls produced no conversation.
How do you prove an AI calling campaign caused the results?
Keep a comparison group that gets no AI call, or a sublist only voice AI can reach. Make the groups similar, for example by random split. Report the gap in goal rate between the groups, not the raw conversion.
Which metrics matter most for AI outbound calling?
Pick one success metric before launch, usually goal outcome rate or cost per goal outcome. Add reached rate and real conversations as support. Add opt-out rate as the guardrail.
How often should you report on an AI outbound campaign?
Send leadership one report at the end of each wave. Give the owner a short weekly status while it runs. For year-round campaigns, use a monthly version with trends.
How do you report AI call compliance and opt-outs to leadership?
Show the opt-out count and rate, how quickly opt-outs were honoured, and the share of sampled transcripts where the AI disclosed itself. Confirm the do-not-call list was applied. Add complaints and any handovers.
Agree the definitions before the first call
Agreed definitions and a control group set in advance make the numbers hard to argue with. Before you start, check one thing: that "reached" and your success metric are signed off in writing.

