The CREATE Framework: Scoring Processes for Automation

A structured way to answer a hard question — which parts of this work should a machine do? CREATE scores each step across six dimensions and lands it in one of four tiers, so the decision is explicit rather than guessed.

CREATE scores a step across six dimensions, then lands it in one of four tiers.

What the CREATE framework is

CREATE is a framework for assessing how ready each step of a process is for automation or AI, by scoring it across six dimensions and assigning it to one of four tiers. It replaces gut-feel “we should automate that” with a structured, step-by-step judgement that's consistent and defensible.

The six dimensions

C — Cognitive complexity: how much judgement, interpretation or nuance the step requires. Higher complexity pushes away from full automation.
R — Repetition: how often and how uniformly the step recurs. High repetition favours automation.
E — Evidence: whether the data and rules the step relies on exist, are reliable and are accessible.
A — Automation alignment: how well the step fits existing systems and tooling.
T — Transformation impact: how much value automating or augmenting the step would actually create.
E — Exception rate: how often the step deviates from the standard case. High exceptions favour keeping a human involved.

Each dimension is scored on a simple scale, and the pattern across the six — not any single score — determines the recommendation.

The four tiers

The scores resolve into one of four tiers for each step:

AGENT — suitable for an AI agent to run end to end.
AUTOMATE — suitable for rules-based or systematic automation.
AUGMENT — keep the person, but give them AI support to work faster or better.
HUMAN — keep it fully human; judgement or exceptions make automation unwise.

Why tiers, not a single score

A single “automation readiness” number hides the reason behind it. The four tiers make the decision explicit — and the difference between AUGMENT and AUTOMATE is often where the real value, and the real risk, sits.

How to use it

Run CREATE across every step of a mapped process and you get a tier for each one — a clear picture of what to automate now, what to augment, and what to leave alone. In LeanBean, this runs automatically as part of the analysis, and its findings feed into the Transformation Coach, which sequences them into a NOW / NEXT / LATER roadmap.

Skip the drawing. Describe it in plain English.

LeanBean turns a plain-language description of any process into a professional map — with waste, efficiency and AI-readiness analysis built in. Free to start.

Build your first map free →

Common questions

What does the CREATE framework stand for?

CREATE stands for Cognitive complexity, Repetition, Evidence, Automation alignment, Transformation impact and Exception rate — six dimensions for scoring how ready a process step is for automation or AI.

What are the four tiers in the CREATE framework?

AGENT (an AI agent can run it end to end), AUTOMATE (rules-based automation), AUGMENT (keep the person but add AI support), and HUMAN (keep it fully human). Each process step is assigned one tier.

How is the CREATE framework used?

Each step of a mapped process is scored across the six dimensions and assigned a tier, producing a clear, step-by-step view of what to automate, what to augment, and what to keep human.