To know if a change on your website actually worked, you need to run it long enough to see a clear winner that isn't just down to luck. This means waiting until the data shows a massive gap between two versions—usually at least 95% certainty—before you decide which one stays. If you stop too early, you're just guessing with your profit.
Why most 'gut feelings' are wrong
I’ve sat in plenty of meetings where someone wants to change a button color or move a phone number because they "reckon" it looks better. Look, I’m all for intuition, but your gut doesn't pay the bills. Data does.
Most small business owners in Brisbane are busy. You don't have time to sit around staring at spreadsheets. You want to know: "If I change this, will more people call me?"
The problem is that numbers jump around. One week you might get ten leads, the next you get two. If you changed your website header on Monday and got a rush of calls on Tuesday, was it the header? Or was it just a busy Tuesday?
This is where the math comes in. We call it statistical significance, but you can just think of it as the "Is this a fluke?" test. If you don't run these tests properly, you're basically flying blind with numbers and making decisions based on noise.
The 'Coin Flip' trap
Imagine you flip a coin ten times. It comes up heads seven times. Does that mean the coin is broken? No. It’s just a small sample. If you flipped it a thousand times, it’d eventually even out to 50/50.
Your website is the same. If 100 people visit your site and 5 buy, that's a 5% result. If you change the text and the next 100 people result in 8 buys, you might think you’re a genius. But honestly? That could easily be a fluke. You need a bigger crowd to be sure.
💡 Quick take: Never trust a result from a test that’s only been running for three days. You need at least one full business cycle—usually a week or two—to account for the fact that people behave differently on weekends than they do on a Tuesday morning.
Step 1: Pick one thing to change
Don't go changing your whole homepage at once. If you change the photo, the headline, and the button color all in one go, and your phone stops ringing, you won't know which change killed your business.
Pick one thing. Maybe it’s the headline. Maybe it’s putting your phone number in big bold text at the top.
We see people do this all the time—they get excited and overhaul everything. Then they wonder why they're wasting cash on marketing that used to work but now doesn't. Stick to one variable. It’s boring, but it’s how you actually get rich.
Step 2: Split your traffic
You need a tool that shows Version A to half your visitors and Version B to the other half. At the same time.
You can't run Version A in January and Version B in February. Why? Because February might be a naturally busier month for your industry. You need a fair fight. Both versions need to go head-to-head in the same weather, the same economy, and the same week.
Step 3: How many people do you need?
This is where most people give up because they think they need millions of visitors. You don't. But you do need enough to make the math work.
If you're a local plumber and you get 500 visitors a month, a 1% difference in results is hard to spot. But a 20% difference? You'll see that pretty quickly.
The rule of thumb is: the smaller the change you're looking for, the more people you need. If you're looking for a massive win, you'll know sooner.
"Angus Smith's take — If you're making decisions based on twenty website visits, you're not marketing, you're gambling with your mortgage money."
— Angus Smith, Founder & Marketing Director
Step 4: The 95% Rule
When you use a testing tool, it’ll give you a percentage of "confidence."
Don't touch anything until that number hits 95%. This means there's only a 5% chance the result happened by accident. In the real world, 95% is about as close to a sure thing as you’re going to get.
If it’s at 70% or 80%, leave it alone. I’ve seen tests look like huge winners at 80% only to completely flip and become losers by the time they hit 95%. Patience is literally money here.
Step 5: Calculate the real-world impact
Let's say Version B won. It got you 2% more enquiries.
Before you celebrate, ask yourself: Is that 2% worth the effort? If you’re doing $10,000 a month, a 2% bump is $200. If it took you ten hours to set up the test, you've barely broken even on your time.
Focus on the big wins. Focus on numbers that pay rent. Changing a button from light blue to slightly darker blue is usually a waste of time for a small business. Changing your offer from "Contact Us" to "Get a Free Quote in 2 Minutes"? That’s where the big money is.
Common mistakes I see in Brisbane
1. Stopping too early: You see a win on day two and turn off the loser. Don't do it. Let it run. 2. Testing things that don't matter: Don't test your font choice. Test your price, your offer, or your main photo. 3. Ignoring the 'Why': If Version B wins, try to figure out why. Did it make people feel safer? Was it clearer? Use that knowledge for your next test.
✅ What to do: Go to your website right now. Find the one thing you think is stopping people from calling you. Set up a simple split test using a tool like Google Optimize (or similar) and commit to not touching it for at least two weeks.
How long will this take?
If you have a decent amount of traffic, you can usually get a result in 2 to 4 weeks. If you’re a very small niche business, it might take two months.
If it’s going to take six months to get a result, you’re probably better off just making an educated guess based on what your customers tell you on the phone and moving on to something else.
What should you do next?
Stop guessing. Every time you make a change to your site because a designer told you it "looks cleaner," you might be flushing leads down the toilet.
Start small. Test your main headline first. It’s the biggest lever you have. If you can get 10% more people to stay on your page just by changing five words, that’s a massive win for your bottom line.
If you want to stop messing about with settings and actually start seeing which of your ads are making money, you need to get your data sorted properly.
Running a business is hard enough without wondering if your website is actually helping or hurting. Get the data, trust the math, and keep the changes that put more cash in the bank.
If you're stuck or you want someone to look over your numbers and tell you if they're rubbish or not, come have a chat with us at Local Marketing Group. We do this stuff every day for businesses just like yours.
Ready to sort it out? Get in touch with us here.