Retention is the money metric

Two companies each acquire 10,000 new users every month, at the same cost. Company L keeps 80% of its base month over month; Company C keeps 95%. That 15-point gap sounds modest. Run it forward and it is the whole story:

• The ceiling is fixed by the leak. A bucket fills until outflow equals inflow: steady-state MAU = monthly new users divided by monthly churn rate. Company L tops out at 10,000 / 0.20 = 50,000 users, forever, no matter how long it keeps spending. Company C's ceiling is 10,000 / 0.05 = 200,000. Same pour, 4x the pond.
• After 12 months Company L sits near 47,000 - already 93% of everything it will ever be. Company C is near 92,000 and still climbing steeply, with years of growth left on the same budget.
• The treadmill test: if next quarter's growth plan is "raise acquisition," ask what the churn rate is first. Pouring faster into a bucket that leaks 20% a month buys a higher plateau, not growth - and the moment spend pauses, the water line falls.

This is why retention is the money metric: it is the multiplier on every acquisition dollar you will ever spend. The numbers above are illustrative arithmetic - in session s2 you will read this exact story off a real cohort chart.

Mobile games make the sequencing brutally clear, because their revenue is voluntary. In free-to-play, only 2-5% of players ever pay anything at all. And the money is astonishingly concentrated: the top 10% of paying players generate roughly 64% of revenue, and the top 1% about 29% (Swrve's payer analysis).

• Whales are not found, they are grown. A big spender is almost never big in week one. High-value players emerge over weeks and months of play - which means they can only emerge from the pool of players who stayed. Kill retention and you have mathematically capped monetization before the first offer is shown.
• The ecommerce echo: the same concentration shows up in shops. Returning customers are typically around 21% of a store's customers but around 44% of its revenue. The repeat buyer is the business.
• The sequencing rule for your roadmap: monetization work multiplies the retained base; it cannot substitute for one. When a team proposes squeezing more revenue from a shrinking base, the arithmetic of Part 1 already told you how that ends.

The documented one. Reichheld & Sasser, Harvard Business Review, 1990 ("Zero Defections"). They measured what happened to per-customer profit as tenure grew, across service industries, and found that cutting customer defections by 5 percentage points raised profits 25-85% depending on the industry (later Bain & Company work stretched the range to 25-95%). The mechanism is boring and durable: long-tenured customers cost less to serve, spend more, and refer. This claim you can quote with a straight face - with its year attached.

The soft one. "It is 5-25x cheaper to retain a customer than to acquire one" traces to a 2014 HBR piece that cites no underlying study. A range that wide is a shrug in numeric clothing. The direction is almost certainly right - acquisition has a real, rising price tag and retention work is often cheaper - but it is folklore wearing a lab coat. Teach your teams to use it as a directional prior, never as a measured constant in a business case.

• LTV:CAC - for every dollar spent acquiring a customer, how many dollars of lifetime value come back; the classic healthy benchmark is 3:1 (David Skok). Retention is the engine of the LTV half: customers who stay longer are worth more, so every retention point raises the ratio without touching ad spend. When to ask: the moment anyone proposes scaling acquisition - "what is LTV:CAC on that channel, and what retention assumption is inside the LTV?"
• Quick ratio - users gained for every user lost: (new + resurrected) / churned. For consumer products, 1.5-2.0 is very good. It is the leaky bucket as a single number: below 1 the bucket is draining even if the top line looks busy. When to ask: whenever MAU is flat and nobody can say why - the quick ratio splits "not filling" from "leaking".
• NRR (net revenue retention) - revenue this period from last period's customers, as a percent of what they paid before. Above 100% means the existing base grows by itself: the business would grow with zero new sales. When to ask: any subscription or B2B review - "if we hired no salespeople next year, does revenue rise or fall?"

Notice what all three have in common: acquisition can flatter each of them for a quarter, but only retention can hold them up for years.