What is AI readiness?
By Joey van Vliet, founderUpdated
In short
AI readiness is how prepared an organisation is to actually put AI to work. You assess it on four fronts: data (is the information available and usable), processes (are they stable enough), people (is there knowledge and ownership) and governance (are the ground rules in place). It is explicitly not about technical infrastructure.
The four dimensions
Data: is the information you need in a system, or only in someone’s head? Is it complete and up to date enough? And are you allowed to use it, technically and legally?
Processes: does the process run largely the same way each time? Is it clear when an outcome is right or wrong? Are there enough examples to test with?
People: is there someone who owns it, not just sponsors it? Do the people who will work with it know what AI can and can’t do? Is there room in people’s diaries, not just in the budget?
Governance: is it agreed which tools are allowed, which data may go into them and who decides? Without those agreements, the first implementation stalls on a privacy question nobody can answer.
Why one score for the whole company says little
A readiness scan that produces a single percentage for the whole company rarely helps. A business can be nowhere near ready for AI in customer service and well prepared for AI in invoice processing, because there the data is structured and the outcome can be checked.
So assess readiness for each candidate process. That’s also what an AI Opportunity Audit does.
Why it matters in practice
Ask four questions per process. If you answer no to two of them, start with a different process.
- Is the input already available digitally somewhere?
- Can someone see within a few seconds whether the output is right?
- Is there someone who really knows this process and has time for it?
- Do we know whether this data may go into an AI tool?
Which processes usually lend themselves well is covered in which business processes are suitable for AI.
Frequently asked questions
Do you need to be AI-ready before you start?
No. Waiting until everything is in order is the surest way never to start. Readiness differs per process: one process can be perfectly suitable while the rest of the organisation isn’t there yet. So assess it per application, not as one score for the whole company.
What is usually the weakest link?
Ownership. Data is often more usable than companies think, and processes are often stable enough. But without someone steering it every week, an implementation grinds to a halt, however well the rest is arranged.
