SAMPLE REPORT — Illustrative company and figures. No live research was performed, actual agent executions: 0, MAX Not run, Guardians not run (built · integration pending). Human decisions remain required.
Sample Report 5 · Customer Retention
Why Are We Losing Customers?
Brightline Facility Services, B2B services (sample) · Sample results only
Client assignment
The company lost 30 of 200 customers last year (15%) and wants to know who is leaving, why, and what retention is worth.
How the client assigns it to TED: The COO shares a sample customer list with start dates, revenue and cancellation notes and asks TED for a churn analysis. TED confirms causes will be treated as hypotheses to test.
Executive summary
Attrition is concentrated in small accounts (21%) while large accounts churn at 6.7%. Lost revenue totals about $807K, mostly mid and large. A targeted program costing about $90K could protect roughly $243K a year, a net gain of about $153K, if the hypotheses hold.
15%
Customer attrition
30 of 200, illustrative
$807K
Annual revenue lost
Sum across segments
$153K
Net retention value
$243K protected − $90K cost
Key findings
1. Small accounts churn most
21 of 100 small accounts left, often within their first year.
2. Mid-size losses cost the most
7 mid accounts at $45K each equal $315K, more than all small losses.
3. Cause hypotheses
Notes suggest slow response times, price comparison at renewal and a missing account review. These are hypotheses until tested.
Attrition by segment (illustrative)
Lost revenue = customers lost × average annual revenue.
SegmentSmall
- Customers
- 100
- Lost
- 21
- Churn
- 21%
- Avg revenue
- $12K
- Lost revenue
- $252K
SegmentMid
- Customers
- 70
- Lost
- 7
- Churn
- 10%
- Avg revenue
- $45K
- Lost revenue
- $315K
SegmentLarge
- Customers
- 30
- Lost
- 2
- Churn
- 6.7%
- Avg revenue
- $120K
- Lost revenue
- $240K
SegmentTotal
- Customers
- 200
- Lost
- 30
- Churn
- 15%
- Avg revenue
- —
- Lost revenue
- $807K
Retention economics (illustrative)
Assumes churn falls to the targets shown.
SegmentSmall
- Churn target
- 21% → 12%
- Customers saved
- 9
- Revenue protected
- $108K
SegmentMid
- Churn target
- 10% → 6%
- Customers saved
- 3
- Revenue protected
- $135K
SegmentTotal
- Churn target
- —
- Customers saved
- 12
- Revenue protected
- $243K (cost $90K)
Options & decision table
Retention options
Illustrative.
OptionQuarterly account reviews for mid/large
- Cost
- $40K
- Focus
- Value proof
Option90-day onboarding program for small
- Cost
- $30K
- Focus
- Early churn
OptionResponse-time commitment
- Cost
- $20K
- Focus
- Service quality
Prioritized action plan
- 1Interview 10 lost customers to test the three hypothesesCOO · 30 days
- 2Start quarterly reviews for all mid and large accountsAccount managers · 60 days
- 3Launch the 90-day onboarding program; win-back offer to recent leaversCustomer success · 60 days
Success metrics
- Small-account churn below 12%
- Mid-account churn below 6%
- Renewal rate tracked monthly by cohort
SOCRATEE's Perspective
Sample text, not generated
Assumptions to test
- Cancellation notes reflect true reasons
- Lost customers are reachable for interviews
What deserves questioning
- Are some small accounts unprofitable to keep?
- Did a competitor or a price change drive renewals?
Sources & evidence
Illustrative assumptions (not evidence)
- All customer counts and revenues are sample data
Verified data needed before deciding
- Billing history by cohort
- Support ticket response times
- Exit interviews
Verified evidence attached: none. No citations are provided because no research was performed.
Safeguards
- Customer contact only by staff, with consent rules followed
- Hypotheses labeled until confirmed
Report status · Research & Review Record
- Actual agent executions: 0
- MAX independent grade: Not run
- 18 Guardians: Built · Integration Pending; no review occurred
- Live research: None performed
- Final decision: Always a human decision
Specialist roles (role definitions · not executed)
- TED — Coordinator: would scope the assignment, route work and assemble the report.
- RES — Would gather customer feedback and service records.
- KNOLLY — Would build cohorts and reconcile billing data.
- SOCRATEE — Would test whether causes are correlation or cause.
- LIBBY — Would check outreach and data-use rules.
- MAX — Potential independent grader. Not run; no grade exists.
- 18 Guardians — Potential oversight layer. Built · Integration Pending; no review occurred.
How AI Team One Works — Step by Step (prewritten)
Stage 1
The assignment
The COO asks why 15% of customers left.
Stage 2
TED coordinates
TED routes cohort analysis, hypotheses and retention economics.
Stage 3
Specialist responsibilities
KNOLLY builds cohorts, RES gathers service records, SOCRATEE separates correlation from cause.
Stage 4
Oversight design
Oversight here would center on keeping causes as hypotheses: SOCRATEE would test whether some small accounts are simply unprofitable to keep, LIBBY would keep win-back contact with staff under agreed consent rules, MAX could grade the cohort and retention math, and the Guardians could review the segment cut-offs. In this sample none of that ran: MAX Not run, Guardians not run (built · integration pending).
Stage 5
The executive report
The result would look like "Why Are We Losing Customers?": findings, options, a prioritized action plan and the data still needed. Prewritten sample content; actual agent executions: 0.