CLV Scenario Simulator

Marketing just proposed a loyalty programme that costs $12 per customer per year. Should you fund it? The answer depends on whether the programme lifts retention enough to offset the cost — and a single-point CLV estimate cannot tell you that. A CLV scenario simulator lets you toggle retention rate, ARPU, and discount rate independently, then compare the resulting lifetime value curves side by side. The input that shifts the curve the most is where you should focus your budget.

Most first attempts get the hierarchy wrong: they tweak ARPU by 10% and celebrate a big CLV jump, ignoring that retention has a compounding effect that dominates over long horizons. A customer who stays one extra year contributes not just that year’s revenue but every subsequent year’s as well. Retention is almost always the most powerful lever.

A sensitivity chart plots CLV on the y-axis against a range of values for one input while holding the others constant. Run it for retention rate first: you will typically see a steep, convex curve. Moving retention from 70% to 80% might add $90 to CLV; moving from 80% to 90% adds $200 — the relationship accelerates because each retained customer feeds into the next period’s base.

Now run the same chart for ARPU. The curve is linear: a 10% ARPU increase produces roughly a 10% CLV increase regardless of starting point. That linearity means ARPU improvements are predictable but less dramatic than retention gains. The discount rate sensitivity is the opposite of retention — higher rates compress future cash flows, flattening the CLV curve. In low-rate environments, long customer lifespans are worth much more; in high-rate environments, early-period revenue matters disproportionately.

A point-estimate CLV of $340 looks actionable until you realise the 80% confidence band runs from $210 to $520. That width is not a weakness of the model — it is the truth about how uncertain the future is. Confidence bands come from varying all inputs simultaneously within their plausible ranges, running many scenarios, and plotting the resulting distribution of CLV values.

Narrow bands mean your inputs are well-constrained — you have solid historical data on retention and ARPU. Wide bands mean one or more inputs are guesses. The practical use: if the lower bound of CLV still exceeds your customer acquisition cost, the investment is defensible even under pessimistic assumptions. If the lower bound dips below CAC, the decision depends on which scenario you believe, and you need better data before committing.

The retention curve plots the fraction of a cohort still active at each period. A constant-rate model (geometric decay) produces a smooth exponential drop: 80% retention means 80%, 64%, 51%, 41%… at periods 1, 2, 3, 4. Real curves rarely look this clean. Most products show steep early churn (the “tourist” users leave fast) followed by a flattening tail (loyal users stick for years).

This shape matters for CLV because a constant-rate model underestimates the value of the loyal tail and overestimates early-period revenue. If your product has a J-shaped retention curve — steep drop, then flat — using a single retention rate will produce a misleading CLV. Better to model two segments: a high-churn segment (60% retention) and a low-churn segment (92% retention), then blend the CLV estimates by segment share.

Ignoring the discount rate. Undiscounted CLV inflates the value of customers who stay for five or ten years. At a 10% annual discount rate, a dollar received in year 5 is worth $0.62 today. Omitting discounting makes long-lived customers look 40–60% more valuable than they are.

Using average retention across segments. If segment A retains at 50% and segment B at 90%, the blended 70% produces a CLV that matches neither group. Segment B is far more valuable, and a blended rate hides that. Always compute CLV per segment, then weight by cohort share.

Projecting indefinitely. A 90% retention rate implies an average lifespan of 10 years. If your product has only existed for two years, projecting ten years of revenue is speculative. Cap the projection at a reasonable horizon (3–5 years for most SaaS products) and treat the remainder as upside, not baseline.

The core formulas behind scenario-based lifetime value estimation:

Scenario: A B2B SaaS product charges $100/month (ARPU = $1,200/year). Monthly retention is 95% (annual retention ≈ 54%). The discount rate is 10% annually. CAC is $800. Gross margin is 75%.

Base case CLV: Using the closed-form, CLV = $1,200 × 0.54 / (1 + 0.10 − 0.54) = $1,200 × 0.54 / 0.56 = $1,157. Margin-adjusted CLV = $1,157 × 0.75 = $868. Payback: cumulative margin exceeds $800 CAC partway through month 12.

Upside scenario: The loyalty programme lifts monthly retention to 96.5% (annual ≈ 65%). CLV jumps to $1,200 × 0.65 / 0.45 = $1,733. Margin-adjusted = $1,300. The $12/customer/year programme cost is trivially covered by the $432 CLV gain. Payback drops to month 9.

Downside scenario: Retention slips to 93% monthly (annual ≈ 43%). CLV falls to $1,200 × 0.43 / 0.67 = $770. Margin-adjusted = $578 — below the $800 CAC. The business loses money on every acquired customer. This is the scenario that makes the retention programme a defensive necessity, not just an upside play.