AI-native revenue forecasting

Turn agency lead data into a revenue forecast you can defend.

Connect your call-tracking data, model the funnel, and project MRR & ARR across conservative, expected and aggressive scenarios — with an AI layer that explains what's driving the numbers and where the risk is.

Where should the forecast pull from?

Choose a revenue-data source. You can switch later.

Step 1 / 4

●Loaded sample agency data — Acme Digital (sample), 12 months.

Leads (last mo.)

—

Avg MoM lead growth

—

Current MRR

—

Top channel

—

Monthly leads — trailing 12 months

Model your funnel economics

Defaults are seeded from your data — adjust to match reality.

Step 2 / 4

Current active clients 26

Your existing book of business (seeded from recent MRR).

Lead → client conversion 6%

Share of new-business leads that become retainer clients.

Avg monthly retainer (MRR / client) $3,000

Recurring revenue per client per month.

One-time setup fee / client $1,500

Onboarding / project fee booked when a client signs.

Monthly lead growth 5%

Expected month-over-month change in new-business leads.

Monthly client churn 4%

Share of active clients lost each month.

Monthly new-business spend $15,000

Optional — enables CAC & payback. Set to 0 to ignore.

Horizon & scenario assumptions

We compute conservative, expected and aggressive in one pass.

Step 3 / 4

Forecast horizon

Seasonality

Apply quarterly seasonality curve

Models typical agency demand swings (slower summer, stronger Q4).

Conservative — growth & conversion factor 0.7×

Aggressive — growth & conversion factor 1.3×

Revenue forecast

Expected scenario · 12-month horizon

Step 4 / 4

MRR forecast — scenario band

Monthly recurring revenue across conservative ↔ aggressive, expected line on top.

Expected MRR Conservative ↔ aggressive band

New clients / month

Signings from converted leads (expected).

Cumulative revenue

Recurring + setup, run-rate to date (expected).

Acquisition funnel

Totals across the horizon (expected).

Lead source mix

From connected data.

✦

AI Insights

Generated locally · deterministic insight engine

Run the forecast to generate insights…

✦ Generate with a real model instead (bring your own API key)

Provider

Model

API key used in your browser only · never stored or sent anywhere but the provider

Exposing provider keys in the browser is fine for a quick demo but never in production — keys belong server-side behind an auth'd route. The architecture doc shows the production AI layer (provider router, streaming, cost tracking, guardrails).

ForecastIQ — a prototype → production demo. Forecasts are modelled estimates, not financial advice. Aurora Dark · GSAP · single-file