How to Run an AI Readiness Assessment (Framework + Example)

The World Economic Forum's Future of Jobs Report 2025 projects that 170 million new roles will be created by 2030 while 92 million are displaced — a structural churn of 22% across the formal workforce. McKinsey research published the same year estimates that AI agents and robotics could technically automate tasks covering up to 44% of US work hours. Those are headline numbers. What they do not tell you is which roles inside your organisation are exposed, which ones are merely changing, and what you need to do about it before the decision is made for you.

An AI readiness assessment answers that question at the role level, not the economy level. Done properly, it takes less than a day of analytical work and produces a ranked action list — which teams to reskill first, which roles to redesign, and where you genuinely have breathing room. This guide walks through the framework step by step, with a worked example you can adapt to your own headcount.

Step 1: Inventory Your Roles

Start with your org chart, not a job-board taxonomy. Pull every distinct job title in the organisation — aim for role categories, not individual names. A 200-person firm might surface 30-40 meaningful role types. For each, note: headcount, department, and whether the role is primarily knowledge-work, physical, relational, or some mix. This inventory is your assessment input; skipping it means you'll produce generic findings that no manager will act on.

A useful filter at this stage: separate roles that produce outputs (reports, decisions, designs, code, data) from those that facilitate process (scheduling, compliance checks, data entry, document routing). Facilitation roles tend to face higher near-term automation exposure. Output roles face more complex augmentation dynamics.

Step 2: Score Task Automatability and Augmentation Potential

For each role, list the five to eight tasks that consume the most time. Then score each task on two dimensions:

• Automatability: Can AI complete this task end-to-end with acceptable accuracy and without human review? Score 1 (no) to 5 (yes, today).
• Augmentation potential: Can AI meaningfully speed up or improve a human doing this task — while the human retains judgment and accountability? Score 1 (minimal) to 5 (high).

Be honest about the augmentation column. Most knowledge-work roles score higher on augmentation than automatability. A senior HR business partner who spends 40% of their time writing first-draft policy documents is not facing replacement — they're facing a 40% reduction in one category of effort, which frees capacity for the relational and strategic work that AI cannot replicate. Conflating augmentation with elimination is the most common error in AI workforce planning, and it produces either panic or false reassurance.

Step 3: Calculate Role-Level Exposure

Weight each task by the share of weekly time it represents, then compute a weighted automatability score for the role. Roles scoring above 3.5 are high-exposure; 2.5 to 3.5 are moderate; below 2.5 are low near-term exposure. Separately, note the augmentation weighted score — this drives your reskilling priority, not just the exposure flag.

A role can be low-exposure and low-augmentation (safe and static), or low-exposure and high-augmentation (safe but opportunity-rich). The combination matters. Do not route all investment to high-exposure roles and neglect the high-augmentation ones — those are where you'll see the fastest productivity dividend.

Step 4: Worked Example — A 12-Person Finance Team

Below is a simplified assessment for a mid-sized finance function. Exposure rating combines automatability score and time-weighted task share. Recommended action follows from both the exposure and augmentation columns.

Role	Automatability (1-5)	Augmentation Potential (1-5)	Exposure Level	Recommended Action
Accounts Payable Clerk	4.6	2.1	High	Redesign role; reskill toward exception handling and vendor management
Management Accountant	2.8	4.3	Moderate	Tool adoption sprint; train on AI-assisted variance analysis and forecasting
Financial Controller	1.9	3.8	Low	Augmentation-first: AI for board-pack drafting; protect strategic judgment time
Treasury Analyst	3.4	3.9	Moderate	Dual-track: automate cash-position reporting; upskill on scenario modelling
FP&A Business Partner	2.1	4.7	Low	High-value augmentation; invest in data literacy and AI-assisted forecasting tools

Notice that the accounts payable clerk — the most automatable role — has low augmentation potential, which means reskilling needs to point toward adjacent roles (vendor relations, procurement operations), not just AI tool training. The FP&A business partner is the inverse: low exposure, high augmentation. That role becomes significantly more valuable with the right tools, and the investment required is modest.

Step 5: Build and Prioritise the Reskilling Roadmap

With exposure and augmentation scores in hand, sort roles into four quadrants:

• High exposure, low augmentation: Role redesign or redeployment. Work with HR to map transferable skills to adjacent, lower-exposure roles.
• High exposure, high augmentation: Urgent reskilling — these roles can survive AI disruption if the person learns to work alongside the tools, but the window is short.
• Low exposure, high augmentation: Productivity investment. These roles will pull ahead of peers at competing organisations if you act now.
• Low exposure, low augmentation: Monitor and revisit in 12 months. AI capabilities are not static.

For each role in the top two quadrants, assign a learning track (tool-specific training, adjacent-role cross-training, or external reskilling programme), a responsible owner, and a 90-day milestone. Roadmaps without milestones are aspirations. Prioritise the roles with the largest headcount first — that is where the aggregate risk and opportunity sits.

One practical note on sequencing: begin with roles where data quality is highest. Assessments built on gut-feel task estimates are less useful than those grounded in time-tracking data, job-analysis surveys, or at minimum a structured conversation with role managers. The quality of your input directly determines the quality of your action list.

How Treeng's AI Readiness Engine Handles This in Under Four Minutes

The framework above is solid and replicable. The friction is the data-gathering and scoring — done manually for a team of 50, it takes the better part of a week. Treeng's AI Readiness engine runs the same role-inventory, task-scoring, exposure-mapping, and roadmap-prioritisation logic against your input, anchored in WEF Future of Jobs 2025 data and McKinsey's occupational-transitions research. It returns a role-by-role exposure map, augmentation priorities, and a sequenced reskilling roadmap, with every finding carrying an evidence grade — solid (strong cross-source consensus), indicative (directionally supported, limited data), or needs data (flag for your own verification). That grading is deliberate: workforce decisions are consequential, and you should know exactly how much weight to put on each finding. The engine is free, permanently.