Publish Less
Most Australian small businesses asking about content marketing should publish less than they're planning to, and fix their service pages first.
That's not a cost-saving argument. It's that a thin service page converts badly no matter how much traffic you send it, and traffic is the expensive part. We've watched businesses publish forty blog posts into a site whose main service page didn't say what the service cost, who it was for, or what happened after you enquired. The posts ranked. The enquiries didn't come.
So before any of what follows: open your main service page and read it as a stranger. If it doesn't answer what you do, who it's for, what it costs, and what happens next, that page is your content strategy for the next two months.
The Realistic Cadence
One properly researched article a month, linked into your existing pages, beats twelve published in a burst and never touched again.
The reason is compounding. An article that earns a few links and a few citations keeps working. Twelve thin ones split the same effort twelve ways and none of them reach the threshold where anything happens.
"Content marketing for small business" runs at 40 searches a month in Australia at difficulty 28, and "how often should I blog" at 10. Low volumes, which tells you something: this is a question business owners agonise over and rarely search. They ask an agency, and the agency usually says "weekly", because weekly is easier to sell than monthly.
One a month. Twelve a year. That's a real body of work by December.
What to Write, Decided in Twenty Minutes
Skip the keyword research for your first ten articles. You already have the list.
Open your sent email. Find the five questions you've answered more than twice this quarter. Each is an article, and you've already written most of it in a reply.
Check your phone notes or your enquiry form. The question people ask before they buy — "do you do X", "how long does Y take", "what's the difference between A and B" — is the highest-converting content you can publish, because the person asking is already most of the way to buying.
Ask whoever answers the phone. They know what people are confused about. They've explained it four hundred times.
That gives you a year of articles in twenty minutes, and every one of them is grounded in something a real customer actually wanted to know. Keyword tools come later, when you've run out of real questions.
The Hour-a-Week Version
If you have one hour a week and no help, here's where it goes.
Week one: talk. Record yourself answering the question out loud, as if to a client. Phone voice memo. Ten minutes, unedited, rambling is fine. Talking produces better explanations than typing does, because you naturally say the thing the client needs rather than the thing that sounds professional.
Week two: shape it. Transcribe it — any transcription tool, or AI — and cut it into headings. Question as the heading, direct answer immediately underneath, detail after. Don't polish yet.
Week three: add the specifics. This is the step that separates your article from the generic version: real numbers, a real example, the thing that surprised you, where this advice doesn't apply. Without this step you've written something that already exists.
Week four: publish and link it. Add it to the site, link it from the related service page, and link the service page from it. Then leave it alone.
Four hours a month. One article. Repeat.
Using AI Without Producing Nothing
AI will write you a competent article about anything in ninety seconds. The output is competent and interchangeable, which is exactly the problem: the same structure and the same claims as every other article on the topic, which is what underperforms.
Where it genuinely helps:
Transcribing and structuring what you said out loud. This is the best use by a distance, because the thinking is yours and the typing isn't.
First drafts of the boring sections. The definitional paragraph, the bit explaining what a term means.
Editing. "Cut this by 30% without losing anything specific" works well.
Finding what you left out. "What would a sceptical reader say is missing here?" produces useful answers.
Where it doesn't:
Generating the opinion. If the article has no position, no number of yours and no example from your own work, it isn't worth publishing regardless of who wrote it. Our rough rule is that at least 30% of an article should be material no general-purpose model could produce.
Google's stated position is that it judges content quality rather than how content was made. In practice the penalty for AI-written content is that it's usually generic, and generic underperforms whoever typed it.
Measuring It Without Giving Up Too Early
Content marketing takes six to twelve months to show in a way that's worth celebrating, which is longer than most people's patience.
Three things worth tracking in the meantime.
Which articles get mentioned on calls. Ask. "I read your piece on X" is the clearest signal there is and no dashboard reports it.
Enquiry rate per article, roughly. You won't get clean attribution; you'll get a pattern.
Branded search volume in Search Console, quarter on quarter. This captures the effect that last-click attribution misses entirely.
What to ignore for the first six months: traffic. Early traffic is dominated by accidental visits from queries you don't want. A piece with 40 visits a month from the right people beats one with 4,000 from the wrong ones. SEO copywriting covers how to write for the right ones.
When to Hire Someone
Three signs it's time.
You've published consistently for six months and it's working. Scaling something that works is a good use of money. Starting something you haven't tested isn't.
The bottleneck is typing, not thinking. If you can explain it clearly in a ten-minute recording, a writer can turn that into publishable work without inventing anything. That's cheap and it preserves what makes it yours.
You've run out of real questions. When the obvious ones are written, keyword research starts earning its keep.
What not to do: hand over the topic choice and the opinions to someone who's never met your customers. That's how businesses end up publishing content that's technically about their industry and recognisably about nothing.
Our content marketing page covers how we work, and the free audit includes a look at whether your service pages are ready for traffic before you go buying any.
Frequently Asked Questions
How often should a small business publish blog posts?
Once a month, properly researched and linked into your existing pages. Consistency compounds; volume doesn't. One article a month that earns links and citations outperforms twelve published in a burst and never revisited, because the twelve split the same effort and none reach the threshold where anything happens.
What should I write about if I have no idea where to start?
Open your sent email and find the five questions you've answered more than twice this quarter. Add the questions people ask before they buy, and whatever your phone-answerer says people are confused about. That's a year of articles, each grounded in something a real customer wanted to know, decided in about twenty minutes.
Is content marketing worth it for a small business?
Only after your service pages work. A thin service page converts badly regardless of how much traffic reaches it, and traffic is the expensive part. Read your main service page as a stranger: if it doesn't say what you do, who it's for, what it costs and what happens next, fix that before publishing anything new.
How long does content marketing take to work?
Six to twelve months before results are worth celebrating. Buying cycles and Google's trust both move slowly. In the meantime, track which articles get mentioned on calls and watch branded search volume quarter on quarter, rather than judging it on traffic, which is dominated early on by visits you don't want.
Can I use AI to write my blog posts?
Use it for transcribing and structuring what you've said out loud, for first drafts of definitional sections, for editing, and for finding what you've left out. Don't use it to generate the opinion. An article with no position, no number of your own and no example from your work isn't worth publishing whoever typed it.
Does AI-written content hurt SEO?
Google judges content quality rather than how it was produced. The practical problem is that AI-written content tends to be generic: the same structure and claims as everything else on the topic. Generic underperforms regardless of authorship. Aim for at least 30% material no general-purpose model could produce.
How much time does content marketing actually take?
About four hours a month for one article, split across four weeks: record yourself answering the question, shape the transcript into headings, add your real numbers and examples, then publish and link it. The recording step matters because talking produces better explanations than typing.
Should I hire a content writer?
When you've published consistently for six months and it's working, when the bottleneck is typing rather than thinking, or when you've exhausted your real customer questions. What not to do is hand topic choice and opinions to someone who has never met your customers. That produces content about your industry and about nothing.
How do I measure content marketing ROI?
Ask callers what they read, watch branded search volume in Search Console quarter on quarter, and look at enquiry patterns per article rather than clean attribution, which you won't get. Ignore traffic for the first six months; early traffic is mostly accidental visits from queries you don't want.
What's the difference between content marketing and SEO?
SEO is making pages findable; content marketing is having something worth finding. They overlap heavily and fail separately: a technically perfect page with nothing to say won't earn links or citations, and a genuinely useful article nobody can crawl won't be read. Most small businesses have the second problem.