82.0% margin · 70.0% credited
Model the lift.
Price the work.
Compare a SaaS baseline with an email-influenced cohort model, then subtract platform, labor, and setup cost before calling the result ROI.
- Monthly cash flow
- Break-even solver
- Local calculation
Credited gross profit exceeds full modeled cost.
$154,713 credited gross profit − $49,650 full program cost across 24 months.
The credited-share input discounts the modeled difference, but it does not replace a holdout, randomized rollout, predeclared attribution rule, or reconciliation with billing data.
$6,450 setup included
Within the 24-month horizon
$19,795 month-end MRR delta
Baseline and email-influenced run rate.
See how lift magnitude changes the decision.
+53.4% ROI · payback M16
+211.6% ROI · payback M8
+292.6% ROI · payback M7
Price the program and isolate the conversion requirement.
Monthly program cost supported after the entered setup cost, using credited gross profit across the horizon.
Additional absolute trial-to-paid lift needed for net contribution to reach zero while retaining the entered churn and expansion effects.
Platform, loaded monthly labor, and other recurring expense.
Do not confuse revenue, gross profit, and net contribution.
Scenario revenue minus baseline before margin or attribution.
Revenue effect after the entered gross-margin assumption.
The share assigned to the email program.
The numerator used in the reported ROI.
Cohort flow, run rate, cost, and cumulative return.
Inputs are internally consistent.
Do not credit the same customer or billing change to both lower churn and higher ARPA.
$105,063 remains after gross margin, attribution credit, recurring expense, and setup investment.
The conservative and upside cases scale entered effects mechanically and do not describe statistical probability.
Export assumptions, cohort flows, scenario sensitivity, cost, attribution, payback, and model boundaries.
ROI requires a baseline that would exist without the proposed change.
The model runs identical monthly customer flows twice. The baseline uses current conversion, churn, and ARPA. The email scenario changes only the entered effects. Incremental value is the difference, not all revenue touched by email.
Opening customers churn, new paid customers enter, and the closing cohort becomes next month’s opening.
Conversion lift is entered in percentage points. Churn reduction is relative to baseline churn. Expansion lift changes scenario ARPA. Because all three effects live in the same recurrence, retained and newly converted customers are not added again as separate lifetime-value windfalls.
Apply gross margin, then subtract labor, platform, operating, and setup cost.
The model uses an average-customer convention within each month and reports month-end MRR separately. It does not treat future recurring revenue as cash received immediately, and it does not label incremental revenue as profit.
The attribution share is an explicit planning discount, not a causal certificate.
Use 100% only when the chosen design supports that claim. Holdouts, randomized rollouts, matched cohorts, billing reconciliation, and predeclared rules can strengthen the estimate. Opens and clicks alone do not establish incremental revenue.
Why use gross profit instead of revenue?+
Revenue ignores the cost of serving additional customers. Gross margin converts modeled revenue into contribution before the email program's own operating cost.
Is conversion lift relative or absolute?+
It is entered as absolute percentage points. Moving from 12% to 14% is a 2-point lift, not a 2% relative lift.
Does churn reduction include dunning?+
It can, but then do not add the same recovered customers elsewhere. Define the entered churn effect as a mutually exclusive net change versus baseline.
Why not multiply new customers by simple LTV?+
Doing that while also projecting monthly cohorts can count the same future revenue twice. The ledger recognizes the modeled customer population month by month.
Is the conservative scenario a confidence interval?+
No. It mechanically applies half the entered effect sizes. It is sensitivity analysis, not a probability statement.