Skip to content
English
Customer portal
Go to RevGravy
  • There are no suggestions because the search field is empty.

3.6 Three-Gate Evaluation Framework

How to decide what to automate next

What this is

Before building any new automation, every idea gets evaluated against three gates. This prevents building automations that nobody uses or that don't produce measurable output. The core principle: the metric is never time saved. It is always output multiplied.


Gate 1 — Capacity (non-negotiable)

Question

Does this remove work from an architect so they can manage more clients — or improve the architect's ability to have meaningful conversations with clients?

G1 is non-negotiable. If an automation does not pass G1, it stops here. This is the only gate that can block an idea entirely.


Gate 2 — Retention

Question

Does this protect client health — through better meeting hygiene, proactive conversations, or improving the client's experience of working with RevGravy?

G2 adds weight toward BUILD NOW. G1 + G2 = strong candidate.


Gate 3 — Compounding Value

Question

Does this get more valuable over time, or does it enable other automations to work better?

G3 pushes BUILD NOW to top priority. Infrastructure automations often pass G3 because future systems depend on them.


Scoring and verdict

Gates passed

Verdict

Action

G1 only

BUILD NEXT

Queue for the following cycle

G1 + G2

BUILD NOW

Add to current or next cycle as priority build

G1 + G2 + G3

BUILD NOW (top priority)

Anchor build for next cycle — start here

No G1

DEFER

Do not build regardless of G2/G3

How to run an evaluation

  1. Write the idea in one sentence: 'When X happens, automatically do Y, resulting in Z'
  2. Run it through G1: does Z remove architect work or improve architect conversations?
  3. Run it through G2: does Z improve client experience or health?
  4. Run it through G3: does Z create compounding value or enable other systems?
  5. Record the verdict in the Automation Catalog (Google Sheet)
  6. Bring BUILD NOW candidates to Phil in the next cycle planning meeting

Core principle

The metric is never time saved. It is always output multiplied. An automation that saves 30 minutes but doesn't increase client capacity or output quality has not earned its place.