Guide
Where to use AI in a business
AI is worth using where a recurring, well-understood piece of work, decision or customer interaction happens often enough that doing it faster or better changes the numbers — and where the information it needs already exists. For most small and mid-sized businesses that means looking in four places: work & operations, decisions & intelligence, customers & growth, and products & services. The right first place is the one with the best mix of value and ease, not the most impressive demo.
Updated
The four places to look
Most AI advice only covers automation. That is one of four areas, and often not the most valuable one.
- Work & operations — repetitive processing, preparation, chasing and filing. Value is usually capacity: hours returned at the rates you pay.
- Decisions & intelligence — recurring decisions and the research that feeds them, such as pricing, quoting, prioritization or forecasting. Value comes from faster or better-informed decisions, with a human still approving.
- Customers & growth — response speed, follow-up, personalization and service. Value may show up as conversion or retention, which should be modelled as a scenario rather than a certainty.
- Products & services — new ways to package expertise. These are exploratory: validate demand before you estimate returns.
How to tell a real opportunity from a plausible one
A real opportunity has four things you can state plainly: how often it happens, what it costs today, whether the inputs are available and structured, and how much judgment the outcome requires. If you can't state them, the next step is measurement, not implementation.
Ask what happens to the value once you capture it. Returned hours only become savings if you reassign, avoid hiring, or grow without adding headcount. Keeping capacity and cash separate stops a business case from quietly double-counting.
A practical sequence
- List 3–5 candidates across at least two of the four areas.
- For each, record frequency, current effort or cost, data readiness and judgment required.
- Rank on value and ease together; start with high value, easier to implement.
- Run a bounded pilot with a named owner, a baseline and a stop condition.
- Expand only after the pilot measures a change against its baseline.
Illustrative example · not a benchmark
Four candidates at a 25-person services firm
| Invoice processing (work) | 312–416 hrs/yr addressable · do first |
|---|---|
| Quote pricing review (decision) | Weekly · human approves · plan next |
| Lead follow-up (customer) | Response-time scenario · validate |
| Packaged advisory product (product) | Exploratory · test demand first |
Figures are illustrative arithmetic, not averages. Your own map uses only the numbers you enter.
When AI is not worth it
- The task happens rarely, so even a perfect tool saves little.
- The inputs are scattered, unstructured or unreliable, and fixing that is the real project.
- Every output needs expert judgment and full review, so little time is actually saved.
- The only case for it is 'competitors are doing AI'.
Compare your own opportunities
The free Opportunity Map ranks what you enter on value and ease, with every assumption shown. About 6 minutes, no signup.
Questions
- What is the best first use of AI for a small business?
- The one with the best mix of value and ease in your business. Often that's a high-volume admin or follow-up task with structured inputs, but it depends on your volumes and systems, so compare a few candidates before choosing.
- Is AI only useful for automation?
- No. Decision support, customer response and new service offerings can matter more than automation. They are measured differently, so compare them as candidates to investigate, not as identical dollar figures.
- How much does it cost to find out where AI fits?
- The WhereAI Opportunity Map is free and takes about 6 minutes. An optional $79 Implementation Planning Brief turns the leading opportunity into a starting plan. It is a planning document, not a quote.