A small business owner — motivated, resourceful, doing everything right by their own standards — opens ChatGPT and types something like: “What’s the best SEO strategy for a local business trying to rank in 2025?”
The response comes back fast. It’s detailed. It sounds authoritative. It covers keywords, meta tags, local citations, content clusters, internal linking. The business owner reads it, nods, screenshots it, maybe pastes some of it into a document they’ll call their “SEO Plan.”
Then they come to us — sometimes weeks later, sometimes months — with a strategy already half-implemented, and a question that quietly signals something has gone wrong: “Why isn’t it working?”
This is not a story about AI being bad. AI is extraordinary. But there’s a version of this story that’s playing out in small businesses everywhere right now, and it’s worth being direct about — because the consequences of getting this wrong are not just wasted time. They can be rankings lost, budgets burned, and in some cases, a website that’s harder to recover than it was before any of this started.
The Tool Is the Same. The Operator Makes All the Difference.
Two people ask ChatGPT the exact same question: “How do I improve my SEO?”
Person A is a business owner who’s never done SEO before. They type the question and hit send.
Person B is a seasoned SEO strategist. Before they type a single word, they’ve fed the AI their current keyword rankings, a competitor gap analysis, their site’s technical audit results, their target audience profile, their content strategy to date, and their business goals for the next 12 months. Then they ask: “Given all of this, where are my biggest opportunities?”
Person A gets a general blog post. Person B gets a performance consultant.
The difference isn’t the AI. The difference is the expertise the operator brought to the conversation — expertise that shaped the input, filtered the output, and knew which parts of the answer were actually applicable to their specific situation.
This is the thing nobody tells small business owners when they’re told that AI is “democratising marketing.” It is, in some ways. But it’s democratising access to information — not access to judgment. And in SEO, PPC, and digital marketing broadly, judgment is most of the job.
What Happens When You Build a Strategy Without Enough Context
We had a client come to us recently with a keyword mapping document they’d built using ChatGPT. They’d done their research. They’d spent real time on it. They’d decided on a pillar page structure, mapped out their supporting content, and were ready to start publishing.
The problem? When we looked at the URL structure AI had suggested, almost every URL contained a variation of the same keyword. The pillar page was /services/seo-for-small-business. The supporting pages were /services/seo-small-business-guide, /services/small-business-seo-tips, /services/small-business-seo-strategy.
To an algorithm, those pages look like duplicates competing against each other. Google sees four pages trying to rank for the same intent and doesn’t know which one to surface — so it often surfaces none of them properly. This is called keyword cannibalisation, and it’s one of the most common SEO issues we clean up. The fix involves restructuring URLs, consolidating content, setting up canonical tags, and in some cases, redirecting pages entirely. That’s not an afternoon’s work.
The client hadn’t done anything wrong by their own logic. They’d asked AI to help with keyword mapping and URL structure. AI had given them what looked like a coherent answer. But AI had no way of knowing that those URL patterns would cannibalise each other — because the client hadn’t told it enough about their site architecture, their existing content, or how search engines evaluate topical clusters. And without that context, AI defaulted to the statistically average response: similar-sounding keywords spread across similar-sounding URLs.
That’s not an AI failure. That’s a context failure. And it’s a category of mistake that costs real money to fix.
The Specific Ways AI Research Can Hurt You — Not Just Fail to Help
There’s a meaningful difference between AI research that doesn’t move the needle and AI research that actively sets you back. Small business owners should know what the second one looks like.
Keyword cannibalisation from incomplete site mapping When you ask AI to suggest keywords without giving it your existing content and URL structure, it has no way to flag overlaps. You end up targeting the same search intent across multiple pages and splitting your ranking potential instead of consolidating it.
Targeting keywords at the wrong stage of the funnel AI will often surface high-volume keywords — because volume looks like success. But a local trades business ranking for a broad informational keyword (“what is HVAC”) gets very different traffic to one ranking for a transactional keyword (“emergency HVAC repair Vancouver Island”). Without understanding your customer’s buying journey, it’s easy to build a content plan that attracts the wrong visitors.
Technical recommendations without site context AI can tell you to fix your page speed, add schema markup, and improve your Core Web Vitals. What it can’t tell you is whether your particular WordPress setup will conflict with the schema plugin it’s suggesting, or whether your hosting environment is the actual bottleneck. Technical SEO advice without site-level context can introduce new problems while solving surface-level ones.
Ad structure decisions without campaign history In PPC, we see business owners restructure their entire campaign based on AI’s “best practice” advice — separating ad groups, changing match types, adjusting bid strategies — without understanding how those changes interact with their specific account’s conversion history. Google’s algorithm uses historical data to optimise. Disrupting campaign structure without understanding that can reset your quality scores and push your CPC up significantly before the account re-stabilises.
AI Is a Multiplier. But You Have to Bring Something for It to Multiply.
This is not an argument against using AI tools in your business. We use them. Most of our clients use them. They save time, surface ideas, and when used well, accelerate work that would otherwise take days.
The argument is this: AI is only as useful as the context and expertise behind the person using it. It doesn’t know your business. It doesn’t know your website’s history. It doesn’t know what your competitors have already done, which tactics are oversaturated in your specific market, or what Google’s most recent core update penalised most in your industry niche.
An experienced SEO strategist or PPC consultant brings all of that to the table — and then uses AI to move faster within a framework built on real knowledge. That’s a very different thing from using AI to build the framework itself.
If you know enough to ask the right questions, AI will give you extraordinary answers. If you’re asking it to tell you what questions to ask, you’ll get something that sounds right but lacks the specificity that separates a strategy that works from one that quietly costs you ground.
What Small Business Owners Should Actually Do
Working with an agency or specialist doesn’t mean handing everything over and hoping for the best. It means bringing professional judgment to the parts of the process where judgment matters most — strategy, architecture, targeting decisions, campaign structure — and then using AI tools within that framework to execute faster and more efficiently.
If you’re at the stage where you’re researching marketing strategies through AI tools, that’s a good instinct. Curiosity is a competitive advantage. But before you act on what you find, ask yourself: does this advice account for my specific website, my market, my customer, and my business goals? If the answer is that you typed a general question and got a general answer, the next step isn’t implementation — it’s a conversation with someone who can apply that information to your actual situation.
The businesses that win in search aren’t the ones who found the best tips on the internet. They’re the ones who built strategies around their specific context — and had the expertise to know the difference between advice that sounds good and advice that actually works for them.
Not Sure If Your Current Strategy Has Gaps? Let’s Find Out Together.
If anything in this post sounded familiar — whether it’s a keyword mapping document you built with AI, a PPC campaign that never quite performed the way you expected, or an SEO plan you’re not 100% confident in — we’re happy to take a look.
We offer a free 30-minute consultation with no pitch, no pressure, and no obligation. Just a straight conversation about where your business is, what you’re trying to achieve, and whether your current digital marketing strategy is actually built to get you there.
In 30 minutes, we can typically tell you:
- Whether your SEO structure has any cannibalisation or architectural issues
- Whether your content is targeting the right stage of the buying journey
- Whether your paid campaigns are set up in a way that’s likely to scale — or quietly draining budget
No forms. No waiting. Just pick a time that works and we’ll talk.
Jyoti Kumari – Founder | Things of Marketing
After nearly a decade working across agencies and in-house teams, She founded Things of Marketing after realising businesses everywhere deserved a marketing partner that was genuinely invested in their growth — someone who asks the hard questions, digs into the data and stays accountable to real outcomes. Not just delivering reports, but helping businesses cut through the noise, make smarter decisions and get results they can actually see in their numbers.