Large companies write AI strategies as long documents. A small business doesn't need that and can't act on it. What it needs is a page that answers five questions clearly enough that you and your team could act on it tomorrow.
What "AI strategy" means for a small business
Strategy here simply means deciding where AI will and won't be used, in what order, with what money and what rules, so that tools are bought to solve problems rather than because a salesperson called.
It's not a technology plan. Most of the page is about your business.
The one-page template
1. The three most expensive problems. Not "we need AI", but problems: "We miss enquiries after 7 pm." "Invoices take two days a month to chase." "Staff spend an hour a day answering the same questions." Put a rough cost or time next to each.
If you did the exercise in masterclass part 1, take these from your "AI can do it" and "AI drafts" columns.
2. The first problem, and what fixed looks like. Pick one. Write the target as a number: "Reply to every enquiry within 10 minutes, including evenings." "Cut no-shows from about 1 in 8 to 1 in 15." If you can't measure it, you can't tell if AI helped.
3. Budget and owner. Setup cost, monthly running cost, and one person responsible, even if that's you. Without an owner, AI projects stall after the launch week. For realistic costs, see how much AI costs a small business in India.
4. Rules. Three or four lines, such as:
- Customer personal data never goes into general AI tools.
- AI may reply to customers on routine questions; refunds, complaints and medical, legal or financial advice go to a person.
- Everything AI sends to customers is logged and reviewed weekly for the first two months.
- Staff use approved tools with their own logins.
For a fuller staff policy, see an AI use policy for your staff.
5. Review date. Every 90 days: is the first project hitting its number? Keep, change or stop it, then pick the next problem from your list.
An example
Problems: (1) WhatsApp enquiries after 7 pm go unanswered until morning; we estimate losing several bookings a week. (2) Payment follow-ups take the accountant about 6 hours a month. (3) New staff take two weeks to learn our services. First: Problem 1. Target: every enquiry answered within 10 minutes, 8 am to 11 pm, with a booking offered. Budget: setup within ₹X; running cost under ₹Y a month. Owner: front-desk manager. Rules: No customer data in personal AI accounts. Assistant handles timings, prices and bookings only. Complaints go to the manager. Weekly log review for 8 weeks. Review: end of February.
The figures are yours to fill in. The point is that each line is specific.
Common mistakes
- Starting with tools. "Should we get ChatGPT Business or Copilot?" comes after you know what problem you're solving.
- Five projects at once. One working project teaches more than five half-built ones.
- No baseline. Write down today's number before you start.
- No owner. "The agency will handle it" is not an owner.
- Ignoring the law. The Digital Personal Data Protection Act, 2023 is law, and its Rules were notified in November 2025. Most duties for businesses (notice, consent, security, breach reporting) apply from May 2027. Build the rules in from the start. See the DPDP guide.
Checklist: your one-page AI strategy
- Three most expensive problems, each with a rough cost or time
- One problem chosen first, with a measurable target
- Today's baseline number recorded
- Budget split into setup and monthly
- One named owner
- Three or four rules on data and human approval
- A review date 90 days out
If you'd like an outside view while drafting it, writing this page with you is how our AI consulting engagements begin. The masterclass ends with a fuller version: your 90-day AI plan.
