The same things turn up over and over. Most of them aren't technical, and all of them are cheaper to fix than they look.
Five problems account for most of what I find when I audit a local business website and its search profiles. Each one below comes with a way to check for it yourself.
1. The tactic was chosen before the goal was defined
"I need SEO" usually means "I'm not getting enough of the right customers." Those are different statements, and the difference is where the money goes.
More walk-in traffic, more phone calls, more high value jobs, more booked appointments: each of those needs different work. Walk-in traffic is largely about your Google Business Profile and whether you show up in the local map results. High value jobs usually come down to what your pages say and how clearly they signal who you're for. Phone calls at specific hours often come down to whether your hours are correct everywhere they appear. The tactics genuinely differ.
How to check: write down the number you want to move and how you would know it moved. If you can't name the number, no amount of search work will tell you whether it worked.
2. The website is treated as finished
A website often gets built once, like a sign bolted to the building, then left while everything around it changes: services, pricing, staff, service area, the questions customers arrive with.
Search engines and AI assistants both favor sources that look current and accurate. A site that hasn't changed in four years collects small errors, and both tend to trust it less. It happens gradually and nothing breaks, which is why it usually goes unnoticed until someone looks.
How to check: read your own services page as if you'd never seen it and count what's no longer accurate. Prices, service area, services you've stopped offering, people who've left. Anything above zero has drifted. That can make a site look neglected, to customers and to search engines.
3. The Google Business Profile is incomplete
It's the most useful free thing a local business has, and it's usually half filled, out of date, or never claimed.
It matters more than the effort it takes, because of what it is: the most structured, most trusted source of factual information about a local business. When Google builds that little map with three businesses listed under it, or an AI assistant answers a question about hours, services, or location, this is the source with the least guesswork in it. That second case is worth understanding on its own, and I've written about how AI assistants decide which businesses to recommend. Gaps don't just fail to help. They can shift the answer toward a competitor whose profile is complete.
How to check: go through every field. Categories, secondary categories, services, service area, attributes, hours including holidays, photos, the description. Then search your own category from a local address and see who appears above you and what their profile has that yours doesn't.
4. Rankings are being measured instead of customers
Ranking first for a phrase nobody searches, or one that attracts people who won't buy, produces a number that improves while the business doesn't.
The strategy usually isn't the problem. The wrong thing is being measured. Rank is easy to see, so it becomes the number everyone watches, and it does tell you something real. It stops telling you anything useful when the phrase is wrong, and a rank report will never mention that.
How to check: for each term you track, ask what someone searching it is trying to do, and whether that's something you sell. Then look at which pages actually produce inquiries and see whether they're the pages those terms point to.
5. AI is being used to produce volume rather than substance
AI tools are genuinely useful for this work. The problem is using them to publish a lot, fast, hoping quantity makes up for not having much to say.
Two reasons it underperforms or doesn't work at all. Search engines have gotten much better at spotting pages that add nothing. Google has said plainly that what it penalizes is writing produced to chase rankings, not writing a machine helped with. More fundamentally, a model doesn't know what your business specifically should be saying: which jobs are profitable, which questions cost you time on the phone, what you do differently. That judgment is yours, and no amount of output replaces it.
How to check: pick a recent page and look for one thing on it a competitor couldn't have published word for word. If you can't find one, there's little reason for the page to be retrieved or cited.
What connects them
Every one of these starts the same way: nobody said what success would look like, so nobody can tell whether the work is getting there. A broken page is obvious. A missing goal isn't.
Which is also why this part rarely gets addressed. It's harder to sell than a report full of metrics, and it produces a shorter list than most people expect to get.
Where this leads
If you want a specific read on which of these is costing you and what to address first, in priority order, that's what a Clarity Review covers. And if the fundamentals are already sound, that's a useful thing to find out too.