Keyword research for small business: why keyword difficulty lies, and how to pick terms you can win
Keyword research for small business comes down to one question: which searches can your site actually win, and are there buyers behind them? For most small sites that means specific, lower-volume phrases, not the big head terms. Keyword difficulty scores help you shortlist, but they can't make the call for you. You have to look at who's already ranking.
I learnt that the hard way doing keyword research for my own site this year. I ran several hundred seed keywords across six topics through Semrush, and the difficulty score was wrong often enough that I stopped trusting it on its own. Here's what it gets wrong, and the process I use now.
What keyword difficulty actually measures
Semrush's keyword difficulty score runs from 0 to 100. It's mostly built from the links and authority of the pages already ranking in the top 10: how many sites link to them, what kind of links they are, and how strong those domains are.
Semrush groups the scores like this:
- 0 to 14: very easy
- 15 to 29: easy
- 30 to 49: possible
- 50 to 69: difficult
- 70 to 84: hard
- 85 to 100: very hard
That's a useful first filter. But notice what it's measuring. It's a summary of other people's backlinks. It doesn't know who those people are, whether their page actually answers the search, or whether the person searching would ever pay you.
Where keyword difficulty lies: three patterns from my own research
A "possible" score with the biggest names in SEO on page one
"Technical SEO checklist" gets about 480 searches a month and scored KD38. On paper, that's "possible". On page one, it's Moz, Ahrefs and Semrush. Those are three of the most linked-to SEO sites in the world, writing about their own subject. A new site isn't outranking them with a better checklist.
"What is a landing page" is the same story at a bigger scale. About 5,400 searches a month at KD39, owned by HubSpot and Unbounce. The score says "possible". The search results say "not this year".
A middling score can hide a page one that's locked up. Big brands often rank on topical authority and brand searches as much as raw link counts, and the score doesn't fully capture that.
A low score you can trust, because page one is weak
It cuts the other way too. When I checked some low-difficulty Melbourne suburb searches, one of the pages ranking on page one was a free Google Sites page. No custom domain, barely any content.
That's the best signal you'll ever get. The score said easy, and the results agreed. If a free page builder is ranking, a proper page that answers the search properly has a real shot.
The score alone can't tell you that. Only looking can.
A score that ignores what the searcher actually wants
Keyword difficulty says nothing about intent, and intent is where a lot of tempting keywords fall over.
- "Electrician public liability insurance" (about 1,000 a month): big volume, and electricians are a group I build websites for. But the person searching wants an insurance quote, not a website. Ranking for it would bring in traffic that never enquires.
- "Car service booking online" (about 20 a month): looks relevant to a mechanic website. It's a driver looking to book a service, not a workshop owner looking for a booking system.
- "Apple Business Connect" (about 880 a month): people searching a product name usually want that product's own site. That's navigational, so it's useful as a heading inside a broader article, not as a page target.
How to pick keywords you can win: the five-step process
- Start with what you sell, not with a tool. List your services and the problems customers have just before they hire you. "Why is my business not on Google Maps" is a problem. "Local SEO" is a service. You want both on the list.
- Pull volume and difficulty in bulk. Paste your list into Semrush's bulk keyword analysis, or Google Keyword Planner if you want something free inside a Google Ads account. You're after a rough sense of demand, not precision. Small numbers are normal in Australia.
- Look at page one for every keyword you shortlist. This is the step most people skip. Search it yourself and look at who's there: national brands and software companies, or small businesses, directories, forum threads and thin pages? If it's the second lot, you can compete.
- Check the intent. Ask whether the person typing this could become a paying customer. If the honest answer is no, drop it, however good the numbers look.
- Give each page one main keyword. Group close variants onto the same page instead of writing a page for each one. Two of your own pages chasing the same search just split the result.
That last step is where research turns into a plan. Mapping every keyword to the page that should rank for it is the core of an SEO strategy built on keywords you can win, and it's the part most worth getting right before anything gets written.
What is a good keyword difficulty for a small business?
For a new or low-authority site, aim for terms scoring under 30, and only keep the ones where page one looks beatable when you check it. Terms in the 30 to 49 range can work if the ranking pages are other small businesses rather than national brands. Leave 50 and above until your site has some links and rankings behind it.
Two from my own list that I've parked for now: "what is a backlink" (about 170 a month, KD72) and "SEO health check" (about 390 a month, KD85). Real demand, wrong fight for a young site.
And the best find on my list was the opposite: "website care plan" at about 320 searches a month and KD6. Decent demand, very low difficulty, and directly tied to something I sell. That's the profile you're looking for.
Low competition keywords: don't write off 10 or 20 searches a month
Most of the keywords that matter to a small service business have tiny reported volumes. That's fine, for two reasons.
Small terms add up. Five phrases about website briefs and checklists each showed 20 to 30 searches a month. On their own, none were worth a page. Together, that's about 110 searches a month on one topic, so they became a single website design brief guide instead of five thin posts.
Zero volume doesn't mean zero value. Every search I tried on industry compliance rules came back at zero. That covered an electrician's REC number, AHPRA advertising rules for dentists and the Tax Practitioners Board disclosure for accountants. I didn't write standalone posts on them. But the facts still sit on the relevant industry pages, because someone deciding whether to hire me will read them, even if nobody types those exact words into Google.
The keyword you're researching right now is a good example. "Keyword research for small business" gets about 70 searches a month at KD11. Small, but winnable, and the people searching it run businesses.
Where local searches fit
If your customers come from nearby, a lot of your best keywords are suburb and service combinations. They often show little or no volume in keyword tools, but they're also where competition is thinnest. That's how a free Google Sites page ends up on page one. Showing up for them takes more than a page on your site. It also takes your Google Business Profile and listings, which is the job of local SEO.
The short version
Use keyword difficulty to shortlist, never to decide. Check page one with your own eyes, drop anything without buyers behind it, and give each page one job. A list of 30 searches you can actually win beats a list of 3,000 you can't.
If you'd rather have someone run the research and build the plan for you, that's part of every SEO package I run. Book a consultation and I'll tell you which searches I think your site can win first.
Frequently asked questions
Is keyword difficulty accurate?
It's a reasonable first filter, but it's built mostly from the backlinks of the pages already ranking. It doesn't account for who those sites are or what the searcher wants. Always check page one before you commit to a keyword.
What is a good keyword difficulty for a new website?
Under 30 is the safest place to start, as long as page one looks beatable when you check it. Scores from 30 to 49 can work when the ranking pages belong to other small businesses. Leave 50 and above until your site has some authority.
How do I find low competition keywords?
Start with your services and your customers' problems, then run the list through a tool like Semrush or Google Keyword Planner. Shortlist the low-difficulty terms, then search each one and look for weak results: directories, forums, thin pages or free site builders on page one.
Are keywords with low search volume worth targeting?
Usually, yes, if the searcher could become a customer. Group related low-volume phrases onto one page so their demand adds up. Some topics with no measurable volume still belong on your service pages, because buyers read them before they get in touch.
Do I need a paid tool for keyword research?
Not to start. Google Keyword Planner is free inside a Google Ads account, and looking at page one costs nothing. Paid tools like Semrush make bulk checks much faster and add difficulty scores, which helps once your list gets long.
Recent Posts
