Keyword Research for Small Australian Businesses: A Practical Method

Start with the job, not the keyword

Before opening any tool, list the three to five jobs you want more of and what each is worth. A plumber who makes $180 on a blocked drain and $9,000 on a bathroom refit should not treat those keywords as equally valuable. This single step is why two businesses with identical rankings can have wildly different returns from search.

Harvest the language your customers already use

Read your last fifty enquiry emails and listen to how people describe the problem on the phone. Customers rarely use industry terms. They search "drain smells" not "sewer diagnostics", and "braces for adults" not "orthodontic appliance therapy". Your own inbox is better raw material than any tool.

Expand, then cut hard

Now use a tool — Semrush, Ahrefs, or even Google's own Keyword Planner — to expand that seed list. Expect it to give you hundreds of phrases. The skill is in the cutting, not the gathering. Anything you cannot imagine a buyer typing on the way to hiring someone goes.

Apply the intent test

For every surviving phrase ask: if someone searched this and landed on my page, is there a realistic path to them becoming a customer within a month? "Emergency electrician Parramatta" passes. "How does electricity work" does not, however high the volume. Informational terms have a role in AI visibility and authority building, but they should not be the backbone of your commercial pages.

Group by page, not by keyword

Search engines and AI assistants both reward one clear page per intent. Cluster your surviving phrases into groups that a single page could genuinely satisfy, then map one page to each group. If two pages would answer the same question, you have created a cannibalisation problem — see why your blog gets traffic but no enquiries for how that plays out.

Check what it costs to win

Look at who currently ranks. If page one is dominated by national brands with thousands of referring domains, a local service business is not going to displace them quickly, and the budget is better spent on Google Ads or on longer-tail suburb terms. Choosing winnable battles is most of the job.

Write it down where people will use it

The output should be a single sheet: keyword group, target page, monthly volume, current position, and estimated value of a customer. If your keyword research lives in a 400-row export nobody opens, it was not research, it was data collection.