Quick answer: Big companies split traffic between two versions; a small site does not have enough visitors for that. Use the one-change-a-month method instead: count this month, change one thing, count next month, compare. Test the button first, then the headline, then how prices are shown, then proof. A change is real when it moves the number two months running.
"A/B testing" sounds like a tool you have to buy. For a site with a few hundred visitors a month, it is a notebook and patience.
The method
- Write down this month's numbers: visitors, taps, messages, bookings.
- Change one thing. One.
- Leave it for a full month.
- Write down next month's numbers.
- If it went up, keep it. If not, put it back. Either way, next change.
What to test, in order
- The WhatsApp button: position, colour, pre-typed message.
- The headline: "Welcome" versus what you do and where.
- Prices: hidden versus starting price versus packages.
- Proof: no reviews versus three next to the button.
- The form: remove it entirely and see what happens.
🔍 Want to know what your website is doing wrong — free? Send us the link. We look at it on a phone, on data, the way a customer does, and send you the three things that are costing you enquiries. Two business days, no obligation.
Get my free check →When a result is real
Small sites bounce around. A jump from 4 to 5 taps is noise; 4 to 12 two months running is real. Seasons matter too — compare November with last November, not with June.
Start
- Pick the month's one change from the list.
- Make it today.
- Set a reminder for the first of next month.
Not sure which step you're on? Book a growth call — fifteen minutes with Lee, no pressure, no obligation. You leave with the step, the first fix and the price.
Frequently asked
Do I need a tool?
No. A sheet with one line per month is the tool.
Can I change two things if I am sure?
Then you will never know which one worked. One.
How many months before I run out of things to test?
You will not. Every month there is a next-best change.