RevOps playbook · AI Employee: Rev

Weekly Forecast Roll-Up

Managers get forecast prep done before the meeting

Install this playbookBack to marketplace
The problem

Forecast day is a synthetic exercise. Reps polish their commit numbers in a spreadsheet an hour before the call, managers roll them up on the drive to work, the VP presents a number nobody trusts, and finance backs into a different number for the board. The signal that would make the forecast honest — stage progression, activity, engagement, the last email the prospect sent — is already in the CRM. It just never gets synthesised in time.

At a glance
Trigger
Cron (weekly)
Approvals
None
What it does
Read-only
Systems
Salesforce · Clari · Slack
How it feels in production

An hour-by-hour walkthrough.

Sunday, 20:00. Rev opens every rep's pipeline in Salesforce. For each open opportunity, it reads: stage, amount, close date, last activity (email, call, meeting), engagement signals from Outreach / Gong / Chili Piper, and the champion's title. It scores each deal 0-1 against a model trained on your own historical win rate — not a generic industry benchmark. For every rep, Rev drafts a commit / best-case / pipeline breakdown with each opportunity's suggested category and the reason ("champion went dark 14 days", "stage-4 for 47 days, average is 21", "three multi-threaded contacts, POC scheduled"). Monday 07:00 the draft lands in the rep's Slack DM. They accept as-is or override with a note (their override is the ground truth — Rev learns from the delta). By 09:00 every rep's forecast is signed off; by 09:30 Rev rolls up to team, region, and segment; by 10:00 the VP sees one dashboard with the numbers, the changes since last week, and the deals driving the delta. Finance sees the same number. The board sees the same number. The forecast conversation shifts from "what's your commit" to "why did DEAL-4211 move from commit to best-case, what do we do about it."
How it works

Step by step.

  1. 01

    Pull every open opportunity + signal

    Read Salesforce (or HubSpot) opportunities. For each, gather: stage, amount, close date, days-in-stage, last activity type + date, contact count, multi-threading depth, engagement from Outreach / Gong.

    Salesforce · HubSpot · Outreach · Gong · Chili Piper
  2. 02

    Score against your own win history

    Not a vendor's generic benchmark. Rev looks at your historical won / lost deals with similar shape (stage, days-in-stage, engagement, industry, ACV band) and computes a fitted probability. Explains the drivers.

    Historical data warehouse · Model
  3. 03

    Draft commit / best-case / pipeline per rep

    Per-rep draft: commit (>75% score, on-track dates), best-case (40-75%), pipeline (rest). Each deal carries the score, the reason, and a suggested next-best-action.

    Reasoning · Slack DM · CRM annotations
  4. 04

    Rep accepts / overrides / annotates

    Sunday-night DM asks for accept-as-is or manual override with reason. Override reasons are captured verbatim; feed the model. Rep's action-plan on at-risk deals goes into the manager's Monday review.

    Slack · Teams · CRM notes
  5. 05

    Roll up + surface deltas

    Team → region → segment → total. Dashboard shows: current commit / best-case / pipeline, delta since last week, top movers (up + down), and the specific deals + reasons driving the change.

    Dashboard · Slack digest · Snowflake reporting mart
Systems and wiring

What you connect to make this run.

Salesforce · HubSpot

read+write

Read opportunities + related activities. Write the forecast category + score + reason back to custom fields so any BI tool sees the same signal.

Outreach · Gong · Chili Piper

read

Engagement signals: cadence responses, call sentiment, meeting attendance. Feed the score, especially the champion-going-dark signal that stage alone misses.

Snowflake · BigQuery

read+write

Historical won / lost + current pipeline for training and re-scoring. Weekly snapshots for board-facing dashboards; deltas by rep, segment, region, competitor.

Slack · Teams

write

Sunday DM to the rep, Monday morning digest to the manager, weekly digest to the VP. Every message deep-links back to the opportunity in the CRM.

What changes

Before and after, honestly.

Time to produce weekly forecast (VP-level)
Before
6-12 hours across reps + managers
After
Under 90 minutes (rep review + manager exceptions)
Forecast accuracy (within 5% of actual)
Before
30-50%
After
70-85%
Deals flagged as at-risk before slipping
Before
20-40% (manager catches by memory)
After
80%+ (score surfaces engagement drop early)
Rep time on forecast prep
Before
60-90 minutes per week
After
10-15 minutes (review + override where needed)
Related

Playbooks that pair with this one.

RevOps

Deal-at-Risk Detection

Sits upstream of forecast; the scoring model is shared.

Frequently asked

Answers about this playbook.

Won't reps just override every deal back to commit?

The override is captured with a reason and then compared to actual outcome. Reps who systematically over-commit see their pattern surface in the manager's dashboard within two months. Every override is a data point, not a hiding place.

How does it handle new reps with no history?

The model uses team-average patterns for new reps until they build enough signal (typically 20-30 deals). Their manager's overrides count double during that ramp period.

What about long-cycle enterprise deals?

The model is stratified by ACV band and cycle length; a $500K deal is not scored against $10K SMB history. For deals over 180 days, the engagement signal weighs more than stage progression.

Can we still use MEDDICC / SCOTSMAN / other qualification frameworks?

Yes — Rev reads whatever custom fields your framework populates and can weigh them explicitly in the score. The framework is the input; the model adds the historical calibration on top.

How does this affect the manager's role?

Managers shift from "chase every rep for a number" to "review the exceptions the model surfaced" — usually 5-10 deals per team per week. More time on coaching, less on data-collection.

See it run on your data.

Free plan, no credit card. Connect the systems this playbook needs and run it against a past event first.

Book a demoMore RevOps playbooks