A four-page Looker Studio diagnostic of 7,043 telco customers: $1.67M a year walking out the door, proof that churn isn’t random, and three costed moves worth ≈ $368k/yr.
// the argument, page by page
Of 7,043 customers, 1,869 have churned, a 26.5% rate worth $1.67M a year in annualized revenue at risk, with a median tenure at churn of just 10 months. Churn is front-loaded: customers in their first six months churn at 52.9% and at 35.9% in months 7–12, falling to 9.5% past four years; more than half of all churn lands inside the first year.
Month-to-month customers churn at 42.7%, 15× the two-year contract rate of 2.8% (one-year sits at 11.3%). Fiber customers leave at 41.9%, twice the DSL rate of 19.0%; customers with no internet add-on churn at just 7.4%. Electronic-check payers are the highest-risk payment group at 45.3%, against 19.1% for mailed checks, 16.7% for bank transfer, and 15.2% for credit cards.
The riskiest cell is new month-to-month customers: 55.2% churn in months 0–6, while two-year contracts hold at 0.0% through their first two years. Fiber churns more partly because fiber customers skew month-to-month; the contract effect holds within every internet type.
Customers without tech support churn at 41.6% versus 15.2% with it; without online security, 41.8% versus 14.6% with. And the effect stacks, with every added protection service lowering churn: 56.7% with none, down step by step to 5.3% with four. This is a perceived-value problem, not just a price problem.
Move 1: convert first-year month-to-month customers to annual with a targeted offer. At a 20% conversion rate, that is ≈ $87K/yr retained (their MRR × conversion × the 42.7% − 11.3% churn gap × 12). Move 2: bundle tech support + security free in year one. At 60% attach, that is ≈ $186K/yr net of an assumed $3/customer/mo service cost. Move 3: a first-90-days onboarding program for new fiber customers, the highest-churn intersection at 69.9%, where a modest 15% relative reduction is ≈ $95k/yr.
The page ends with a what-if slider: drag the month-to-month conversion rate and watch retained revenue recompute from $116K of first-year M2M MRR. The business case is interactive, not a static claim.
Look at the anatomy of each move: a trigger (a customer enters a segment), a rule (contract type, tenure, attach status), a message (offer, bundle, onboarding email), and a log to measure it by. That is, node for node, the shape of my n8n work: the invoice pipeline is Gmail + a rules gate + Sheets, and Move 3’s onboarding drip is the audit harness’s loop-and-throttle pattern with a different payload. This dashboard doesn’t end in slides; it ends in an automation backlog.
The loop runs the other way too. The audit harness applies this dashboard’s discipline (ground truth, a controlled run, a measured gap) to the automation itself. Analysis writes the automation’s next backlog; automation writes the analysis’s next dataset. That’s the whole site in one sentence.