- The budget most small businesses have
- Start with what's already free before you spend a penny
- Where it's worth paying, and what it costs
- A real example: the £120 a month experiment
- Step by step: your first 30 days
- The uncomfortable part nobody puts in these guides
- Where AI earns its keep beyond content
- The mistake that wastes the most money
- When to stop DIY-ing it
- Frequently asked questions
- Sources worth reading
The short version: you can run a effective small business marketing engine on AI tools that cost under £50 a month, most of it free, if you pick two or three tools and use them instead of collecting subscriptions you never open. The tools were never the bottleneck. Your offer, your consistency, and your willingness to edit what the AI hands you are the bottleneck, and that's the part nobody selling you a "10 AI tools every small business needs" listicle wants to say out loud.
The budget most small businesses have
When I say "budget marketing" I don't mean a vague notion of being frugal. I mean the real number I hear from small business owners week after week: somewhere between £0 and £150 a month for marketing tools, on top of whatever time they can personally give it, which is usually five to seven hours a week squeezed around running the business.
That's not a sad number. That's a workable number if you spend it right. The mistake I see constantly is business owners signing up for four or five AI tools in one weekend of enthusiasm, using none of them past week two, and concluding "AI doesn't work for businesses like mine." It's not that AI doesn't work. It's that a scattered toolkit never gets deep enough use to pay for itself.
Start with what's already free before you spend a penny
Before you buy anything, here's what costs nothing and does real work:
- ChatGPT free tier (GPT-4o mini access): drafts social captions, email subject lines, first drafts of blog posts, customer FAQ answers.
- Google Gemini free tier: good for research and pulling together competitor summaries, connects to your Gmail and Docs if you're in the Google ecosystem already.
- Canva free plan: Magic Write and basic Magic Studio image tools are included, enough for social graphics without paying for Pro.
- Meta AI, built into Instagram and Facebook business tools: free caption and image suggestions inside the app you're already posting from.
- CapCut free version: auto captions, background removal, basic editing for short video, which matters because video is where most of the free organic reach still is.
If you're only using free tools well, you're already ahead of most small businesses I meet, because most aren't using any of them consistently, let alone all five.
Where it's worth paying, and what it costs
Once you've squeezed the free tier and hit its limits, here's where I'd spend real money, with real 2026 pricing:
- ChatGPT Plus, around $20 a month (about £16), gets you longer memory, file uploads, and the ability to build a custom GPT trained on your brand voice, your FAQs, your tone of voice document. This is the single highest-value upgrade on this list.
- Canva Pro, around £10.99 a month, adds brand kits so every graphic pulls your exact fonts and colours automatically, plus background remover and resize magic that saves real hours.
- Buffer, free for up to three channels, then around £5 a month per channel after that: schedules posts so you're not manually posting daily.
- Descript, around £12 a month, edits video by editing the transcript, which matters if you're turning one piece of video into five clips a week.
That's a realistic full stack for under £45 a month. Compare that to what an agency retainer costs for the same output, usually £800 to £2,000 a month for a junior account manager posting on your behalf, and the maths on doing it yourself with AI support is obvious.
A real example: the £120 a month experiment
I ran this with a client last year, a small bathroom and kitchen fitting business in Kent, three staff, no marketing person, a marketing budget of £120 a month that had previously gone nowhere on boosted Facebook posts.
We cut the boosted posts entirely. Instead: ChatGPT Plus (£16) to draft weekly customer story posts and email newsletters based on quick voice notes the owner sent me after finishing jobs, Canva Pro (£11) for before-and-after graphics using their brand kit, and CapCut, free, to turn phone footage of finished bathrooms into 30-second reels with auto captions. Total spend: £27 on tools, the rest of the £120 went on a small monthly boost of the one post each month that performed best organically, rather than boosting everything blind.
In four months, their Facebook page went from roughly 40 enquiries a quarter to 71, and two of those became jobs worth £8,400 and £11,200. The tools didn't do that. The consistency did. What changed was that posts went out every week instead of whenever the owner remembered, because the AI removed the "I don't know what to write" excuse that had been stopping them for two years.
Step by step: your first 30 days
Here's the actual sequence I'd follow if I were starting from nothing:
- Week 1: Write one paragraph describing your brand voice (three adjectives, one thing you never say, one phrase you always use). Paste it into ChatGPT and ask it to remember this for every future request. This single step fixes 80% of the "AI content sounds fake" problem people complain about.
- Week 1: Set up Canva's brand kit with your logo, fonts, and colours so every graphic is on-brand without thinking about it.
- Week 2: Pick one platform where your customers are, not the one you personally enjoy. If you're a local service business, that's usually Facebook or Google Business Profile, not TikTok. Check what gets seen; timing matters more than people think, and it's worth reading up on the best time to post on Facebook for small business pages before you set a schedule you'll never adjust.
- Week 2: Batch-write four weeks of content in one sitting using ChatGPT, feeding it real details (a customer question, a job you finished, a mistake you fixed) rather than asking it to "write a post about our services," which produces generic mush every time.
- Week 3: Schedule everything in Buffer's free plan so posting doesn't depend on you remembering at 4pm on a Tuesday.
- Week 3: Turn one piece of existing video, even old phone footage, into three short clips using CapCut's free auto-caption tool.
- Week 4: Look at what got engagement and do more of exactly that. Not more content generally. More of the specific thing that worked.
The uncomfortable part nobody puts in these guides
Here's the bit that gets left out of most "AI for small business" content: AI does not save you time on marketing. It saves you time on the first draft. If you are not spending nearly as long editing what the AI gives you as you used to spend writing it from scratch, your content is going to read exactly like everyone else's AI content, and your customers can tell. I see it constantly, the same rhythm, the same "In today's fast-paced world" openers, the same three-item lists with an em-dash between each. It is not that AI writing is bad. It's that unedited AI writing is instantly recognisable and it makes a small business look exactly as generic as the big ones it's trying to compete with.
The businesses getting real results are the ones using AI to get past the blank page faster, then spending real time putting their own specific details back in: the customer's actual name, the actual number of years they've been trading, the actual mistake they made in 2019 that taught them something. That editing time is non-negotiable, and if you skip it to save the last twenty minutes, you've wasted the whole exercise.
Where AI earns its keep beyond content
Content gets all the attention but it's not where AI does its best budget work. A few underused applications:
- Research: before writing anything, ask ChatGPT or Perplexity to summarise what your top three competitors are posting about and what tone they use. There's a legal line here worth knowing, so it's worth understanding what you can legally learn about a competitor's marketing before you start copying their strategy wholesale.
- Idea generation when you're stuck: rather than staring at a blank content calendar, use it to generate new angles and business ideas that aren't the same ten suggestions every business in your sector already tried.
- Customer service drafts: feeding your FAQs into a custom GPT so it drafts email replies to common questions, which you then check and send, cutting reply time from ten minutes to two.
- Video, which is expensive to outsource: there's a growing set of useful free AI video editor tools small businesses should know about, and this is the area where the free tier gap between AI-assisted and professionally edited has closed the fastest.
The mistake that wastes the most money
The single most expensive mistake I see isn't overspending on tools. It's building content around a channel with no audience instead of building an owned home for your content first. Too many small businesses pour AI-generated content into social platforms they don't own, where an algorithm change can wipe out their reach overnight, and they have no email list, no website content, nothing that's theirs. Before you scale up your AI output, it's worth thinking about how to build a content creation website for your business so the work you're doing compounds somewhere you control, rather than evaporating into a feed. Platforms like TikTok have useful growth programmes worth understanding if you're trying to build an audience there, but treat it as a discovery channel, not your only home.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
When to stop DIY-ing it
There's a point where the DIY approach hits a ceiling, usually around the time you're spending more than ten hours a week on marketing and it's still not converting, or you've got the budget growing past £500 a month and no clear plan for it. That's the point where paying for a few hours of proper setup, someone building your custom GPTs correctly, setting up your automation, training you on the actual workflow, pays for itself faster than another month of trial and error. I've written before about what that kind of AI consultant support costs, and for most small businesses it's a one-off few hundred pounds, not an ongoing retainer.
Related reading: It's Not About the Money.
Related guides live in the AI for Small Business: 25 Plain-English Guides to Automation, Marketing, Sales and Admin.
Related: ai consultant cost what you get.
Related: how much time ai consultant saves ten person business.
Related: growth marketing: guidelines and how to pitch.
Related: business strategy (this site).
Frequently asked questions
How much should a small business spend on AI marketing tools per month?
Somewhere between £0 and £50 a month is realistic for most small businesses, covering ChatGPT Plus, Canva Pro, and a free scheduling tool like Buffer; spending more than that without a clear content plan almost never improves results.
Which free AI tool should a small business start with first?
Start with ChatGPT's free tier and give it a written brand voice paragraph before anything else, since that single step fixes most of the generic-sounding output people complain about.
Can AI replace hiring a marketing person for a small business?
It replaces the drafting and admin work, not the strategy and judgement, so it works well for a business owner doing marketing themselves but doesn't remove the need for someone deciding what to say and to whom.
How do you stop AI-generated marketing content sounding fake or generic?
Spend as long editing the AI's draft as you would have spent writing it yourself, adding real specific details like names, numbers, and dates the AI couldn't have known, since unedited AI content follows recognisable patterns customers now notice.