Most Brisbane business owners I talk to are obsessed with what I call 'The Big Lie.' They open their dashboards, see that total revenue is up 10% from last month, and pat themselves on the back.
I hate to be the bearer of bad news, but that number is almost entirely useless.
If you’re running a cafe in West End or a tradie business in Logan, looking at your 'total monthly users' is like trying to judge the health of a forest by looking at a photo of the canopy from a plane. It looks green, sure. But you have no idea which trees are dying, which ones are new saplings, and which ones have been there for twenty years.
This is where cohort analysis comes in. It sounds like academic jargon, but it’s actually the most honest way to look at your data. If you want to know if your business is actually growing or if you’re just on a treadmill of expensive customer acquisition, you need to stop looking at totals and start looking at groups.
What is Cohort Analysis (Without the Fluff)?
In plain English, a cohort is just a group of people who share a common characteristic over a specific period.
Instead of looking at all your customers as one giant, messy bucket, you slice them up. The most common way to do this is by 'Acquisition Date.' For example, everyone who made their first purchase from your Milton-based e-commerce store in January 2024 is the 'January Cohort.'
Cohort analysis is simply the act of watching that specific group over time. Do they come back in February? Are they still buying in June? Or did they all disappear the moment your 'New Year' discount ended?
I’ve seen this backfire more times than I can count: a business spends $10,000 on a massive Meta ads campaign in March. Their total sales spike. They think they’re geniuses. But when we look at the cohort data three months later, we realise that 95% of those March customers never returned. They weren't customers; they were one-night stands that cost the business $50 each to acquire.
The Three Cohorts That Actually Matter
Don't let some 'data scientist' convince you that you need 50 different segments. For most Australian SMEs, you only need to master three types of cohorts to see real growth.
1. The Time-Based Cohort (The 'Loyalty' Check)
This is the bread and butter. You group people by when they first interacted with you.The Scenario: You run a boutique gym in Newstead. You want to know if your '6-Week Challenge' actually builds long-term members or just 'resolution hunters.' By tracking the January 1st cohort vs. the March 1st cohort, you can see if your retention drops off a cliff after the initial challenge ends. If the March cohort (who joined without a discount) stays longer than the January cohort, your marketing shouldn't be focused on New Year's resolutions—it should be focused on whatever you did in March.
2. The Channel-Based Cohort (The 'Quality' Check)
Not all leads are created equal. This is where most agencies completely miss the mark. They’ll tell you Google Ads are 'performing' because the Cost Per Lead is low.But if you look at the cohort of people who came from Google Ads vs. the cohort of people who came from your organic LinkedIn posts, you might find that the LinkedIn crowd spends 3x more over their lifetime. This is why measuring ROI is so difficult when you only look at the first click. Cohort analysis tells you which channels bring in 'stayers' and which ones bring in 'strays.'
3. The Behaviour-Based Cohort (The 'Trigger' Check)
This is for the overachievers. You group people by an action they took.The Scenario: An e-commerce brand we worked with in Fortitude Valley realized that customers who bought a specific 'Starter Kit' were 400% more likely to buy again within 90 days compared to those who just bought a single item. By identifying this behaviour-based cohort, they stopped wasting money on general ads and poured everything into pushing that Starter Kit.
Why Your Current Setup is Probably Lying to You
Here’s a frustrating truth: Google Analytics 4 (GA4) is a powerhouse, but its default reports are often garbage for this kind of thinking. Most people just look at the 'Real-time' or 'Acquisition' tabs.
Look, I get it—another article telling you to 'dive into your data' is maddening when you're busy running a business. But if you're relying on the standard 'out of the box' reports, your GA4 setup is junk. It aggregates everything, which hides the very patterns you need to see.
You need to go into the 'Explorations' tab in GA4 and specifically build a Cohort Exploration. It’s not as hard as it looks, but it requires you to stop looking at the 'Total' row at the bottom of the spreadsheet.
The 'Leaky Bucket' Realisation
We learned this the hard way back in 2019 with a client in the tradie space. They were spending more and more on marketing every month, and while revenue was growing, their profit was flat.
When we ran a cohort analysis, the problem was glaring. Their 'Customer Retention' for any cohort older than three months was effectively zero. They were pouring water into a bucket with a massive hole in the bottom.
Instead of buying more ads (which is what every other agency suggested), we told them to stop. We spent three months fixing their follow-up process and customer service. We didn't need more 'new' people; we needed the people from the 'May Cohort' to actually call them back in August.
This is why I argue that benchmarks are bullshit. Who cares if the industry average retention is 20%? If your cohorts are showing a downward trend month-over-month, you have a fundamental business problem that no amount of 'optimised' ad spend will fix.
How to Run Your First Cohort Analysis This Week
You don't need a degree in statistics. You can do this in Excel or Google Sheets if you have your sales data.
1. Export your sales data: You need Customer ID, Date of First Purchase, and Date of Subsequent Purchases. 2. Create your 'Month 0': Group your customers by their first purchase month. 3. Track the 'X Months Later' Spend: For each group, calculate how many of them spent money in Month 1, Month 2, Month 3, and so on. 4. Look for the 'Cliff': Does every group drop off at Month 4? That’s your cliff. That’s where you need to send a 'We Miss You' email or a special offer.
The Contrarian View: When Cohort Analysis Fails
I’ll be the first to admit that cohort analysis isn't a magic wand. There is an exception here: if you have a very low volume of sales (e.g., you sell three high-end yachts a year), cohort analysis is a waste of time. Your sample size is too small to mean anything.
But for 90% of Brisbane SMEs—retailers, service providers, SaaS, or hospitality—this is the only way to see the truth.
Stop building dashboards nobody uses that just show pretty line graphs of 'Total Sessions.' Start asking: "What did the people we acquired in July do in October?"
That single question will tell you more about the future of your business than any 'Marketing Report' your current agency is sending you.
Summary Checklist for Business Owners
Ignore the 'Total' numbers: They hide the churn. Identify your 'Cliff': Find out exactly when your customers stop coming back. Compare acquisition channels: Stop spending money where the 'one-hit wonders' come from. Fix the bucket first: If your older cohorts are disappearing, stop spending on new ones until you know why.
Marketing in 2026 isn't about who can shout the loudest; it's about who understands their customer lifecycle the best. Cohort analysis is the map that gets you there.
Tired of reports that don't tell the full story? At Local Marketing Group, we cut through the vanity metrics to find what actually drives profit for your Brisbane business. Let’s talk about your data.