This article expands on the diagnosis behind the Applied AI Test.
TL;DR
- AI readiness isn't a single parameter — it's seven dimensions, from data quality to executive sponsorship.
- The weakest dimension usually decides whether your first implementation succeeds — not the strongest one.
- Most mid-sized companies are "ready enough" to start with one process — provided they know where their gaps are.
- The point of the diagnosis isn't a grade — it's a choice: which process to implement first, and what to watch.
Why "are we ready for AI" is the wrong question
The question "are we ready for AI" suggests a yes/no answer. In practice, readiness is multidimensional: a company can have excellent data but zero executive backing — or an engaged team, but processes so disorganized there's nothing worth automating. The first implementation usually trips over the weakest dimension, not a lack of technology.
So instead of asking "whether", it's worth asking "where are we strong and where do we have a gap" — and starting with a process that doesn't depend on your critical gap.
The seven dimensions of readiness
1. Data
Does the information AI would use exist in a usable, readable form? This isn't about "big data" — it's about whether documentation, spec sheets, or order history are available and reasonably consistent.
2. Processes
Is the process you want to improve repeatable and describable? AI supports processes with a clear input and output. A process based purely on "gut feel" needs to be named first.
3. People and skills
Is there someone on the team who will own the solution on the business side? Not an AI expert — a process expert who can tell you whether the output is actually good.
4. Technology and infrastructure
Can your systems (ERP, drives, tools) be safely connected? Usually less is needed than you'd think — a first micro-app rarely requires rebuilding your stack.
5. Security and compliance
Do you know which data is sensitive and where it can be processed? This is the dimension that's easy to ignore at the start and painful to discover later.
Does someone on the board genuinely want this to succeed, and will they free up the team's time for it? Implementations without an owner on the decision-making side go quiet after the first week.
7. A culture of experimentation
Does the company tolerate a version 0.1 that works "well enough", refined through iteration? Expecting a finished product from day one kills most pilots.
Example: A company with excellent data and zero executive sponsorship will lose to a company with mediocre data but a CEO who asks about progress every week. The weakest dimension wins.
How to use this diagnosis in practice
Score each dimension on a simple scale (e.g. 0–3) and find your two weakest. That's not a reason to hold off — it's a signal for where to start and what to watch. If data is your weakest dimension, pick a first process where the data is clean. If sponsorship is weakest, start with a process that delivers a fast, visible result to build trust.
Practical tip: Don't try to shore up all seven dimensions at once before starting. Choose your first process so it sidesteps your critical gaps, and catch up on the rest along the way.
What's next?
Want this diagnosis calculated instead of eyeballed? The Applied AI Test scores all seven dimensions based on your answers and returns a PDF report with your weakest points and a recommended first process — in ten minutes.
