Free up hours, cut costs, and boost consistency. AI isn’t the future for small businesses. It’s now.
Thanks mostly to the launch of ChatGPT three years ago, the term AI is everywhere. AI isn’t new, but the proliferation of new capabilities is. The past few years have seen a gush of new generative Artificial Intelligence (AI) applications that can fundamentally transform how some of our work gets done. Generative AI is a type of artificial intelligence that creates new content, including text, images, audio, video, or even computer programming code, based on patterns it has learned from vast amounts of existing data. Instead of simply retrieving or classifying information, it uses machine learning models, most often large language models (LLMs) or diffusion models, to predict what comes next in a sequence.
This allows it to “generate” outputs that can be original in form while still reflecting the style, structure, and context of its training data and the input prompt. Common examples include chatbots that write articles or answer questions, image generators that produce realistic or artistic visuals from a text prompt, and tools that compose music or design prototypes.
There are thousands of gen AI tools available now, and names such as ChatGPT, Gemini, CoPilot, and Claude are entering the vernacular. Normally offered as a web app, making them accessible and ubiquitous, these tools are helping shave hours off administrative tasks, improve response times, and make marketing materials more consistently reflect a coherent brand voice.
While generative AI can dramatically speed up creative and analytical tasks, it is still in its infancy, and sometimes says things that reveal its tender age. Its results still require human review for accuracy, ethics, and alignment with intended goals.
In short, AI can be wrong and sometimes hallucinates! Use it blindly at your peril!
Not all AI is Generative AI
We won’t dwell on it here because for now, gen AI is where it’s at for small business owners, but there are many other types of AI that don’t create new content. It’s all over the place. AI engines recommend your next online purchase or Netflix movie. Predictive analytics helps foresee and avoid equipment failures. Classification models are used by your email provider to sort emails into “spam” or “not spam,” and flag fraudulent transactions to your credit card companies. Delivery and logistics companies use optimization algorithms to route trucks for the shortest travel time or most energy efficiency.
In the future, AI tools will help physicians and other clinicians diagnose and even predict disease or disorders in patients. Signals will be controlled in real time so route traffic can optimally routed in areas where congestion is a problem. The number of potential use cases is unlimited and while some small businesses should pursue these sorts of AI tools in their operations, we will concern ourselves here with generative AI.
What Generative AI Actually Does (Plain English)
Think of it as pattern recognition, plus autocomplete.
It studies piles of examples, spots patterns, and predicts what likely comes next. From that, it can summarize long content into key points, draft emails and posts, classify and tag information, recommend next steps, and surface similar items. Hooked up to your apps, it can fetch facts, create tasks, fill forms, and nudge you with reminders so routine work moves faster and stays organized. Myriad forms of output are possible including text, images, ppt files, spreadsheets, computer code and audio/video.
It’s best at clear, well-structured tasks where examples exist. Give it context, a goal, and a tone, and it will produce a first pass you can refine. It handles variations quickly (multiple subject lines, product blurbs, or code snippets) and it keeps format and style consistent. It’s also useful for extracting data from text, summarizing meetings, and proposing next-best actions based on past patterns. With basic tool access, it can update a calendar invite, draft a follow-up, or log notes in your system.
In most cases today it remains an assistant, not a replacement. You set intent, provide constraints, and make the call; it does the groundwork and iterations. Treat it like a hyper-capable junior: very fast, with an abundance of energy, and occasionally wrong. Very wrong. These are the jobs that will be replaced by AI. Junior positions.
Here are some real use cases of my own that will help to drive this home.
Early Startup Research
Earlier this year, I visited Della Terra in St. Catharines, ON, a specialty shop that sells mostly olive oils and balsamic vinegars. It’s a beautiful little store with very good quality products and a terrific vibe. The concept was new to me and I was intrigued by it, so when I got home I put this prompt into ChatGPT:
Have a look around Southern Ontario at every town larger than 25,000 people and see whether or not they have an olive oil shop. Tell me which ones do and which ones don’t. Give me a summary of the current operators in the province regardless of community size.
There’s a lot going on here. The model was smart enough to add the criteria to rule out stockists or those who merely carry the products of others along with a large variety of other products, thus not qualifying as a specialty olive oil shop.
A few prompts later and I had a detailed listing of all the competition in the province, and the locations that looked promising for a new operation.
I knew their address, website, and basic offerings.
Summarizing Feedback
Each year, I conduct an end-of-year exercise with our graduating entrepreneurship students, in which they provide recommendations on what activities we should Stop, Start, and Continue. We use a word cloud tool (Menti) that allows students to submit words or phrases, which are then displayed and updated on the screen in real time.
I then pick and choose themes from the word cloud for discussion during the class.
Although the submissions can be exported to a spreadsheet, the class size of 100 students leads to significant duplication, with numerous recurring themes across three documents (Stop, Start, Continue), each containing approximately 100 responses. Additionally, there is overlap between the documents, as certain issues may be positioned as either a start or stop.
The last time we did this I submitted the three excel docs to ChatGPT and asked it to analyze and summarize them, eliminating duplicates and grouping the comments into common themes. The setup took 1 minute, the response took 1 second, and this all saved me at least an hour.
But this is what I call “monkey work”. It doesn’t take smarts. It just takes time (for humans, that is). So “monkey workers” should probably be worried about their jobs.
Helping with “Blank Page Syndrome”
Blank page syndrome is real. That sinking feeling when you’re supposed to create something smart and useful, but all you’ve got is an empty screen. It happens to everyone—even experienced writers.
This is where generative AI shines. Instead of sitting stuck, you can toss it a handful of keywords, a rough outline, or even just an idea that isn’t fully formed yet. In seconds, you’ve got a draft to react to.
Whether you’re drafting a business proposal, newsletter, or report, AI helps you get moving and lowers the barrier to creativity.
I hope you’re not surprised that I leaned on ChatGPT and Perplexity in writing and researching this very piece.
Marketing & Content, Quicker Campaigns, Consistent Output
Marketing is a great place to start looking for opportunities to leverage AI tools. Here’s how you might use AI to draft your next newsletter and create three social media posts from it in under 30 minutes. Imagine you have the following objectives:
- Generate newsletter outlines, subject lines, and calls to action tailored to your audience.
- Turn one long post or video into platform-ready snippets and captions.
- Research and identify SEO keyword clusters, draft articles, and produce meta titles/descriptions.
- Brainstorm ad angles, headlines, and copy, then analyze performance summaries.
- Create simple visuals, thumbnails, lightweight graphics, using AI design tools.
Here’s a potential prompt for an email campaign:
You are a small-business email marketer. Draft a 200-word newsletter for [audience], promoting [offer] with a friendly, helpful tone. Provide 5 subject lines (max 45 characters) and 3 preview lines. Include one clear CTA.
Similarly here’s a potential prompt you could use for drafting a SEO blog:
Create a detailed outline for a 1,000-word blog post about [topic] aimed at [audience]. Include H2/H3s, keyword clusters, 6 FAQs, and suggested internal links based on [list your pages].
HR Policies
You aren’t big enough to have a dedicated HR person. Unfortunately, that’s you! But you’ve come to the point that you think you should have some policies around hours, pay, leave, etc. so that you and your management make decisions fairly and consistently.
You have Google Gemini create the first draft with this prompt:
You are an HR consultant for a 15-person small business. Create a clear, simple set of HR policies covering hours of work, overtime, vacation/leave, pay periods, statutory holidays, sick leave, workplace conduct and ANY OTHER AREAS YOU THINK ARE APPROPRIATE. The policies should be written in plain language (9th grade reading level), no legal jargon, and formatted as a handbook-style document with headings and bullet points. Include a short introduction about why policies matter for fairness and consistency. Flag any areas that may require review with a local employment lawyer or government labor standards. Keep the draft concise.
Gen AI is still a Baby
I laugh when I hear about how AI is going to replace all of our jobs.
It’s almost a ritual: each new wave of technology arrives, surrounded by anxious predictions that it will render us all obsolete. Do you know what else was supposed to replace us?
- The cotton gin
- The computer
- The personal computer
- The internet
- Robots
Each of these, at the margin, did indeed eliminate jobs some jobs, at the margin. But taken as a whole they are multiplicative. They make a society more productive. Think about where we’d be without the internet. It’s almost unthinkable. The benefits it has brought to society are immeasurable, but not without casualties.
- How would you like to be a travel agent right now?
- Or a magazine publisher?
- Record store?
- Phone directory ad salesperson?
AI won’t be any different. There will be a combination of winners and losers, with most of us being winners, one hopes.
Don’t lose sight of the fact that AI is still a baby. These tools will literally make up facts. Like any tool, used without care, it can be dangerous. A combination of AI with a lack of critical thinking could lead to embarrassing and costly mistakes. Check everything!
Also, be careful about pasting sensitive info into unvetted tools. User agreements universally favour those writing them, and as a general rule, I would avoid use of any personal data with these tools.
Not sure where to start? Ask ChatGPT (Or Perplexity, or Gemini, or Copilot, or Genspark)
Let’s get real meta and ask AI how you should use AI in your small business. Use something like the following prompt in any of the popular LLMs and see what comes out:
“I am the owner of a small business in the [industry] sector, located in [country/region]. My business has about [number] employees and our main challenges are [list top 2–3 pain points, e.g., marketing consistency, customer service load, inventory management]. Act as a consultant who specializes in applying AI for small businesses. Suggest practical, low-cost ways I can use AI tools to improve efficiency, save time, or increase revenue in my business. Please give me ideas across different areas such as marketing, sales, operations, finance, and HR. For each idea, explain the benefit, suggest the type of AI tool to use, and give one simple example prompt I could try.”
Every successful small business owner that I know is a lifelong learner. AI is just another opportunity for you to set yourself apart and use your critical faculties to find ways to leverage these powerful tools, while avoiding the rabbit holes that many go down.
Start small. Measure everything. Keep what works.
If you liked this, you might like this related post about systems thinking. How can you use AI to help you adopt this?
