How-to guides

Workflows · Revenue

A revenue forecast that updates itself when the pipeline moves.

Spreadsheet forecasts die because nobody updates them on Friday. Claude in the loop reads the pipeline daily, computes what changed, and tells you when the forecast actually moved.

Why the default doesn't work

Founder-stage revenue forecasting is structurally broken. The spreadsheet was right last Wednesday. Today it's wrong because a deal slipped, a renewal got pulled in, and a stage moved. Nobody updates it because updating it is annoying and the cost of being wrong is hidden until the board meeting.

The 'just use the CRM forecast' answer also fails, because the CRM forecast is downstream of next-step text and close dates that nobody maintains. Garbage in, garbage out, every quarter.

Claude with the pipeline wired changes the equation. Daily reads of the deal record, automated detection of moved/slipped/stalled deals, and a synthesis layer that says 'forecast moved 12% this week, here's why.' The honest forecast.

What changes when an agent runs the workflow

Daily computed forecast

Every morning: weighted pipeline value by stage, close-date adjustments, momentum signal. The forecast is a derived quantity from real deal state, not a hand-edited number.

Slip detection

Deals where the close date moved out, stages that went backward, deals where activity dropped. The agent flags them daily. The 'we'll close in Q3' that quietly became 'we'll close in Q4' surfaces the day it slips.

Conversion priors that learn

Your actual stage-to-stage conversion rates inform the forecast math. Stage 3 to closed-won at 18%? The forecast reflects that, not a textbook 30%.

Honest variance reporting

'Forecast moved $X this week, here's the deals that drove it.' Investor updates, board updates, and your own week-in-review get the actual delta instead of a guess.

Three steps to running it

1

Connect the pipeline

Salesforce, HubSpot, Pipedrive, Notion, or spreadsheet. The agent reads deals: stage, probability, value, close-date, last-touched. No new system to maintain; whatever you have now works.

2

Calibrate stage priors

First two weeks: the agent learns your stage-to-stage conversion from history. After that the forecast uses your priors, not generic SaaS priors. The math reflects your actual funnel.

3

Read the daily delta

Forecast snapshot lands in your morning brief. 'Q3 forecast +$48K (deal X moved to closed-won), -$25K (deal Y slipped to Q4).' You see the moves without a Friday review meeting.

What people running this say

The forecast finally tells me the truth on Wednesday, not on the Sunday before the board meeting.

Founder, vertical SaaS (paraphrased)

I caught two slipped deals in week one that nobody had flagged. The Q3 number was off by $40K and I would not have known.

Head of Sales, fintech

Forecasts are how you find out you have a problem. Make the forecast cheap to keep current and you find problems faster.

Patrick Hillstrom, Nerve

Common questions

Is this a replacement for Clari or Gong Forecast?

No, those are enterprise-grade RevOps platforms. Nerve is the founder-and-early-stage version: lighter, faster to wire up, and tuned for teams that don't have a RevOps function yet. If you're already on Clari, keep it.

How does it handle one-off deals or unusual stages?

Custom stages and probabilities are honored. The agent uses your conversion priors per stage, not a fixed waterfall. One-off deals with manual close-date overrides work fine.

What about renewals?

Renewals and expansion are first-class. The agent tracks them separately from new logo. The forecast splits new ARR from renewal ARR so you can see each lever moving.

Can it forecast cash, not just bookings?

Yes, if you give it payment terms. The agent translates closed-won + terms into cash-arrival timing. Useful for runway-conscious founders.

How accurate is it in practice?

Better than spreadsheet hygiene allows. The accuracy gain is mostly because the data is current, not because the math is fancy. Daily updates beat weekly stale data by a lot.

Roles where this matters most

Stop guessing on Sunday night. Forecast on autopilot.

Start free, no credit card

Other workflows