Choosing keywords that bring buyers
Buy the searches of people who are ready to act, and refuse the rest. That means starting with a short list of phrase and exact match terms that describe what you sell, checking the search terms report every week to see what people really typed, and adding negatives for everything that was never going to buy.
Intent matters more than volume
A term with a thousand searches a month can be worth less than one with forty, if the forty are people with a wallet out. Sort candidate terms by what the person wants, not how many of them there are.
| Search | What they want | Worth paying for |
|---|---|---|
| emergency plumber near me | Someone right now | Yes |
| plumber prices melbourne | To compare and then hire | Usually |
| how to fix a leaking tap | To do it themselves | Rarely |
| plumber jobs | A job with you | No |
Match types in plain terms
- Exact — shows for that search and very close versions of it. Tightest control, least volume. Write it as [emergency plumber].
- Phrase — shows when the meaning of your term is contained in the search. A sensible default. Write it as "emergency plumber".
- Broad — shows for anything Google judges related. Most volume, most waste. Only use it once you have solid conversion tracking and a long negative list.
Start with phrase and exact. Broad match can be useful later as a discovery tool, in its own ad group, with a tight budget.
Negatives are half the job
Negative keywords stop your ad showing for searches you never want. They are the cheapest improvement available to most accounts, and nobody adds enough of them.
- Free, cheap, DIY, how to, salary, jobs, courses — almost always negatives.
- Competitor and brand names you do not sell.
- Places you do not service, even if your targeting should have handled it.
- Products adjacent to yours that you do not stock.
Read the search terms report weekly
Keywords are what you bought. Search terms are what people actually typed. The gap between the two is where the money leaks.
- 01Open the search terms for the last seven days.
- 02Sort by cost, highest first.
- 03For each expensive term with no conversion, decide: negative it, or is it fine and just unlucky?
- 04For each term that converted and is not already a keyword, add it as an exact match keyword.
- 05Ten minutes a week is enough.
Keep a list worth reusing
The terms you learn are an asset. Keep them somewhere — the ones that convert, the long-tail phrases worth testing, the negatives you always apply, and the words competitors clearly bid on. Atlas has a keyword library for exactly this, so a new campaign starts from what you already know rather than a blank page.
Common questions
- How many keywords do I need to start?
- Ten to thirty across a few ad groups is plenty for a small business. Depth beats breadth early on.
- Should I bid on my competitors' names?
- You can, and it is legal in Australia to bid on the name as a keyword, though using it in ad text is another matter. Expect a high cost per click and a low conversion rate; test it small.
- Is broad match bad?
- Not bad, but expensive to learn on. It needs working conversion tracking and a long negative list before it earns its place.
- What are long-tail keywords?
- Longer, more specific searches — fewer people, but clearer intent and usually a lower cost per click. They are often the best value for a small budget.