How do small businesses actually use AI?

Small businesses mostly use AI for everyday writing and admin rather than anything futuristic: drafting emails, proposals and content, replying faster to enquiries, researching prospects, summarising calls and handling routine back-office tasks. The ones who get real value pick a few measurable tasks, give the AI their own business facts to work from, and keep a person checking anything that reaches a customer.

Recommended resource: for the full step-by-step method, The AI Advantage by Manthan Bhendkar (founder of Asendify) covers this in depth: The Playbook for Using AI to Move Faster and Win Customers. PDF, £304.50, instant download.

This page gives a broad picture: what the research says AI is good and bad at, the common uses by type of business, and the guardrails that keep it from embarrassing you. If you want to know where to start, see which tasks to automate first.

What the research says AI actually does well

Several well-known studies point the same way:

• Customer support: in a study of more than 5,000 support agents by Brynjolfsson, Li and Raymond, an AI assistant that suggested replies raised issues resolved per hour by 14% on average, and by 34% for the newest, least-skilled agents (NBER working paper).
• Business writing: Noy and Zhang found professionals using ChatGPT on realistic writing tasks took 40% less time and produced work rated 18% higher in quality (Science, 2023).
• Consulting work: in the Harvard Business School and BCG study, consultants using AI on suitable tasks finished 12.2% more tasks, 25.1% faster. On a task deliberately outside AI's ability, they were 19% less likely to get it right (HBS working paper).

Read together: gains on the right work are large, AI lifts less experienced people the most, and on the wrong work it makes people confidently worse. The researchers called this a "jagged frontier". For a small business, the lesson is to choose tasks carefully rather than use AI for everything.

The most common uses, by type of business

These are the places AI most often pays off first:

• Agencies and freelancers: prospect research and personalised outreach, proposal drafts, client reports, turning call notes into tasks, repurposing client content.
• Consultancies and professional services: meeting summaries and follow-ups, first drafts of reports, research briefs, searching across past work.
• Local service businesses (trades, clinics, salons, property): instant replies to enquiries, quote drafts, appointment reminders, review requests, answering the same ten questions.
• E-commerce: product descriptions from specs, customer service replies, analysing reviews and return reasons, abandoned-basket and post-purchase emails.

Across all of them, the back office is a quiet winner: inbox triage, invoice reminders, pulling data from receipts and forms, writing standard operating procedures from a screen recording, and a weekly summary of leads, sales and overdue items.

How customers feel the difference

Customers don't care whether you use AI; they care what they get. Where AI helps a small business stand out:

• Speed: a reply in minutes and a proposal the same day, when competitors take a week. (See how to reply to leads faster.)
• Consistency: the same accurate answer every time, from a single source of business facts.
• Availability: routine questions answered outside office hours, with a clear route to a human. (See cheap AI customer support.)
• More included: hours saved on repetitive work can go into service or keep prices competitive.

A worked example: a two-person plumbing business

A small plumbing firm misses enquiries while on jobs. It sets up three things over a month:

1. An instant, honest reply to web enquiries that confirms receipt, asks for the postcode and photos, and says when a person will call back.
2. Quote drafts built from the job notes and a fixed price list, which the owner checks and sends.
3. An automatic review request sent a few days after a finished job.

Each system has one number to watch: response time, quotes sent per week, and reviews received. If a system doesn't move its number, it gets switched off.

The guardrails that matter

Small businesses that get burned by AI usually skipped one of these:

• AI is never the source of a fact. Prices, policies and claims come from your own documents, not the model's guesses.
• Nothing reaches a customer unchecked until a system has earned trust over time.
• No fake reviews, testimonials, clients or results. The FTC's 2024 final rule explicitly covers AI-generated fake reviews.
• Respect data protection. In the UK, the ICO publishes guidance on AI and data protection.
• Measure it. If you can't measure a system in hours, speed or money, it's a hobby.

Key takeaways

• Most small-business AI use is practical: writing, replies, research, summaries and admin.
• Research shows big gains on suitable tasks and worse results on unsuitable ones.
• The best first uses depend on your type of business; customer-facing speed often wins most.
• Feed AI your own facts, keep a human check, and give every system a number to track.
• Never let AI invent proof, prices or promises.

FAQ

Is AI worth it for a very small business?

Often, if you pick one or two tasks you do every week and can measure. Start small, measure the time saved, and stop anything that doesn't pay for itself.

Do small businesses need to pay for AI tools?

Free plans are enough to test ideas. For customer data and regular use, a paid business plan is usually better because of data controls and limits. Check each tool's terms.

Will AI replace staff in small businesses?

In practice, it more often removes repetitive tasks so the same people can do more valuable work. The research above shows the largest gains for less experienced staff.

What should a small business never use AI for?

Anything where a wrong answer would be costly and unchecked: final prices, legal or medical claims, and invented social proof.

For a step-by-step version of all this, from an opportunity audit to a 30-day plan, see The AI Advantage; the free AI time savings calculator estimates the hours a task could give back.

More playbooks on other topics: https://www.asendify.co/books/

Looking for a different topic? Browse all 23 Asendify books by topic.