How to Assess AI Readiness in Your Organisation
How to identify opportunities for AI to enhance processes, improve products and develop new offerings within your business.
The term AI gets used so loosely that it has almost stopped meaning anything. It is more useful to treat it as a set of tools, some mature, some emerging, each suited to specific problems in specific contexts. And it should be used to improve the human experience of work, not just the speed of it.
This guide is about figuring out where you actually stand. Not every organisation is ready to act on AI, and starting without clarity on your capabilities and constraints wastes time and money. The three exercises below will help you identify what to tackle first and, just as importantly, where to hold off.
Sustainability is built into the process rather than treated as a separate concern. The environmental and social implications of your technology choices belong in these early decisions, not in a review six months later.
Exercise 1: The AI Readiness Assessment
AI projects that begin with solution thinking before the problem is well defined tend to stall. This exercise cuts through that instinct by evaluating five critical readiness areas.
Rate each area from 1 (not at all) to 5 (very much so):
| Area | Question | Score (1 to 5) |
|---|---|---|
| Data Availability | Do you have accessible, organised data relevant to your business challenges? | |
| Tech Infrastructure | Can your current systems support additional software integrations? | |
| Team Skills | Does your team have basic data literacy or experience with analytical tools? | |
| Leadership Support | Are decision-makers open to data-driven approaches and new technologies? | |
| Process Mapping | Are your workflows and processes well-documented? |
If you want to work through these exercises as a team, we have put together a downloadable worksheet that covers all three.
As you assess your systems and data, it is also worth asking whether you are aware of the energy use behind your digital tools. They all have a footprint, and understanding this early helps you make choices that balance performance with impact.
Once you have scored each area, total your score and assess against:
- 0-10: Focus on foundational data practices before AI implementation
- 11-17: Ready for targeted, limited-scope AI solutions
- 18-25: Well-positioned for broader AI initiatives
A low score does not mean failure. It may reflect that your priorities lie elsewhere, or that conditions are not yet right. If that is the case, the most valuable next step may not involve AI at all. Focus on strengthening data quality, simplifying systems, or resolving operational bottlenecks. These improvements are often overlooked but will create far better conditions for AI adoption if and when the time is right.
Exercise 2: Identifying Specific Business Problems for AI
Broad goals like “improve efficiency” or “reduce admin” are too vague to guide AI adoption. You need to identify one to three concrete problems where:
- The problem is clearly defined and well understood
- Relevant data exists or can be collected without major disruption
- Solving the problem would create tangible business value
A few examples to provoke thinking across functions:
| Area | Common Pain Point | AI Application |
|---|---|---|
| Customer Service | Repetitive queries | FAQ chatbot |
| Document Processing | Manual data entry from forms and invoices | Automation tools |
| Inventory Management | Over- or understocking | Demand forecasting |
| Marketing | Inconsistent engagement | Content personalisation |
| Sales | Poor lead prioritisation | Lead scoring model |
Document your top three priorities. If you list more, treat them as backlog candidates, not active workstreams.
Not every efficiency gain is worth pursuing. Before selecting a use case, consider whether solving it with AI is more effective than solving it with a simpler process change. AI should not be the default solution.
Sustainability check
Looking for a double win? Try identifying areas where AI can reduce both cost and environmental impact, whether that’s cutting unnecessary print, reducing redundant processing, or automating low-value, high-energy tasks.
Exercise 3: Setting Measurable Objectives
For each identified use case, define success in concrete terms. Abstract benefits like “streamline operations” are not sufficient.
Apply the SMART framework to define your objectives:
- Specific: Clearly define what improvement looks like
- Measurable: Include quantifiable metrics
- Achievable: Realistic given your current capabilities
- Relevant: Aligned with broader business goals
- Time-bound: Include timeline expectations
Example objectives:
- Cut invoice processing time from 3 days to same-day for a wholesale operation handling 200+ invoices per week
- Reduce failed deliveries by 25% over the next quarter by adding postcode validation and address matching
- Improve stock forecasting accuracy by 15% over the next year, reducing overordering and waste
- Halve the manual document handling and associated printing across customer onboarding within 3 months
Be honest about whether the metric is a proxy for impact or impact itself. For example, reducing “response time” is only valuable if it correlates with customer satisfaction or cost reduction. Otherwise, it is a vanity measurement.
What to do with this
If your score in Exercise 1 is low, do not skip ahead. Foundational gaps in data or systems will undermine even the best pilot, and the exercises above will have shown you where those gaps are.