How the forecaster computes its number

Revenue model overview

The forecaster partitions year-1 WhatsApp revenue into four streams. Each stream answers a different “what would not have happened without WhatsApp?” question, and the streams are defined to be mutually exclusive so they sum to a single number without double-counting any one purchase.

• Welcome flow — purchases that happen on the first session, attributed to the welcome sequence.
• Cart recovery — only the incremental uplift over the brand’s existing email-based recovery, never the full recovered cart.
• Reorder uplift — the second purchase a welcome-flow buyer wouldn’t have made without the WhatsApp nudge, capped at one per cohort within the first year.
• Personalized campaigns — purchases from a small number of personalized monthly messages (capped at three), each running through a read → engagement → conversion funnel.

Monthly cohort engine

Time dynamics matter. A flat annual rate would imply revenue arrives uniformly across the year, which doesn’t match how a WhatsApp list actually grows. The engine builds a per-month cohort of new opt-ins, ages each cohort forward, and applies churn to the cumulative stock so the numbers respect the fact that subscribers decay.

Each channel’s rate is a two-stage funnel: first, a fraction of the channel’s volume clicks through to WhatsApp (scans the QR, taps the link, etc.); second, a constant 90% of those click-throughs actually opt into the list. The per-industry click-through rates are tabulated below; the 90% opt-in is a global constant.

Channels split into two families. Recurring channels (online sessions and in-store walkthroughs) contribute the same monthly opt-in pool every month — they refresh organically. Finite stocks (email list, SMS subscribers, social followers) opt in once and then taper off; the engine spreads each channel’s annual opt-in pool across the year with a 60.00% / 25.00% / 10.00% / 5.00% decay tail (M1 / M2 / M3 / M4+ flat).

A constant 2.00% monthly churn is applied to the cumulative subscriber stock. This is a deliberate model commitment rather than a slider.

The four stream formulas

welcome_revenue[m] = new_optins[m] × welcomeCvr × AOV

recovery_revenue[m] = subscribers[m] × monthlyPurchaseAttemptRate × cartAbandonRate × incrementalWaRecoveryUplift × AOV

reorder_revenue[m] = welcome_buyers[m − reorderLagMonths] × repeatUplift × AOV

personalized_revenue[m] = eligible_subscribers[m] × messages_per_month × readRate × engagementRate × conversionRate × AOV

Personalized-campaign funnel constants (global): 80% read · 30% engagement · 15% conversion.

Per-industry priors

Every number the engine uses, by industry. The five “→ WhatsApp” columns are the per-channel click-through rates before the global 90% opt-in stage is applied. Pick your row and check the forecaster’s number against numbers you already trust.