Not every business is ready for AI. Rolling it out where it cannot stick wastes money, burns goodwill with the team, and makes the second attempt twice as hard. Use this check before you spend anything.
Readiness has nothing to do with being technical. It comes down to whether the work is repeatable and whether the data touches a system.
Five signs you are ready
1. The same reports get rebuilt every week
If someone assembles the same sales, stock or margin report every Monday, that is a standing appointment AI can keep instead. Recurring work is the easiest work to hand over.
2. The work lives in a few core systems
An ERP, a web store, an accounting package, a 3PL portal. If the data is already digital, AI can read it, move it and summarise it. It does not need to be tidy, only digital.
3. Your best people do data entry
The planner you hired for their judgement spends half their week re-keying. That is a setup problem rather than a staffing problem, and it is the fastest return you will find anywhere in the business.
4. The answers live in one person's head
"Ask Sarah" is the knowledge base. When the answers to common questions actually live in systems and inboxes, AI can serve them to everyone, and Sarah gets her day back.
5. You need to grow without hiring
Orders are up and the headcount budget is not. If the plan is to do more with the same team, the manual work has to go somewhere.
Three signs you are not ready yet
None of these is permanent. Each one needs fixing before AI will pay for itself, and pretending otherwise is how rollouts fail.
1. The core process changes every week
If nobody can describe how the work is done because it is different every time, fix the process first.
2. The data lives on paper and in heads
AI cannot read the notebook in the warehouse office. If key information never touches a system, get it into one first. That step is smaller than most people fear, but it has to come first.
3. Nobody owns it yet
Buying AI and hoping for the best rarely works. A rollout needs an owner, a first department and a definition of done.
What the score means
Three or more of the ready signs and none of the blockers: start with one department, prove it, and expand from the results. Most teams see measurable time savings within the first few weeks, and the early wins usually cover their own cost before the rest of the work ships.
Hit a blocker: fix that first and keep the scope small. A process worth automating usually needs about a fortnight of tidying first. Working out what to fix, what to automate and in what order is what the AI Ops Audit settles.
The common mistake is buying licences while the answer is still "not sure".