Conversion · Conversion rate

What should count as a conversion for my business?

Pick the one customer action closest to revenue as your "main conversion." For a business that takes bookings, it's a completed booking; for a tutoring academy where consultation comes first, it's a consultation request.

Count the actions leading up to it separately as "supporting actions." If supporting actions are high but main conversions are low, customers are stopping somewhere in between, which makes it easier to find what to fix.

When choosing a conversion, also consider whether you can actually count it. If bookings go through an outside service and you can't count the completion screen, start by treating booking button clicks as a temporary conversion and comparing them with your own booking records.

When your business changes, change your main conversion too. If an academy that used consultation requests as its main conversion opens online payments, completed payments become the main conversion from then on.

How do I measure conversions?

Set up an event so your analytics tool knows when a customer completes the action, then mark that event as a key event in GA4. GA4 calls these "key events" (shown as "major events" in the Korean interface), while ad tools like Google Ads still call them "conversions."

For a small business, one main conversion and one or two supporting actions are enough as key events. Setting too many actually blurs which numbers matter.

Always note the date you change settings. If numbers suddenly jump or drop from that day, it may be the measurement that changed, not customer behavior.

1. 01Decide the moment your main conversion is complete. Example: when the booking confirmation screen opens, or when the contact form's submit button is pressed.
2. 02Set up an event to be recorded at that moment. For inquiries and consultations, you can use generate_lead, Google's recommended event name.
3. 03Mark that event as a key event in the event list in GA4's Admin.
4. 04Make a booking or inquiry yourself once and check that it shows up in the Realtime report.
5. 05If you run ads, connect the same action as a conversion in your ad platform too.
6. 06Note the setup date, and use the numbers from the following 2–4 weeks as your baseline.

How do I count conversions that happen off my website, like phone calls or KakaoTalk inquiries?

On your website, you can only count how many times the call button was tapped. Analytics tools can't tell whether the call actually went through or whether the customer booked afterward.

So treat button clicks as a sign of interest, and match actual bookings against your own records. If you add a column to your booking log for how customers found you, such as website, map, Instagram or word of mouth, you'll see a rough pattern within a few weeks.

Another approach: a craft studio can note on a website-only class offer, "Mention that you saw us on our website." The numbers won't be perfect, but they help your decisions far more than knowing nothing.

If you take inquiries through a KakaoTalk Channel, check the stats in the channel's admin screen too. Put them side by side with your website's KakaoTalk button clicks, and you'll get a sense of how many customers who tapped the button actually started a chat.

How do I calculate what one conversion is worth?

If you roughly work out how much revenue one conversion brings on average, it's easier to decide how much you can spend on ads or website improvements. A ballpark figure based on your own records is enough; it doesn't need to be exact.

For example, if a Pilates studio's records show that 3 out of 10 trial bookings turn into full memberships, and each membership brings in a certain average payment, then "one trial booking = membership revenue × 0.3." These numbers are only an example; the rate differs for every business, so always calculate it from your own records.

Use this value to compare against your ad cost per result. If the ad spend to get one trial booking is higher than this value, it's time to revisit your ad targeting or landing page.

This calculation is useful even if you don't run ads. It helps you estimate how many conversions it takes to earn back what you spend on website improvements or a photo shoot.

What should I fix first to get more conversions?

First find where customers stop, then fix the earliest problem first. Put your supporting action and main conversion numbers side by side, and you'll see where people drop off.

If you change several things at once, you can't tell what worked. For a hair salon, move the price list next to the booking button this month and cut down the booking form fields next month, changing one thing at a time.

Asking customers directly is another good way to choose what to fix. A single question to a customer who came in for a consultation, like "Was there anything hard to find on our website?", can give you an answer faster than the numbers.

After making a fix, compare periods of the same length. Match them, such as 4 weeks before and 4 weeks after, and note any closures or events in between so you're less likely to misread the results.

• The first screen loads quickly on a phone and the booking and contact buttons are visible right away
• Information needed to decide, such as price, location and hours, is close to the booking button
• The booking form doesn't ask for information you don't really need
• Reasons to trust you are visible, such as reviews, work photos and an introduction of the owner or director
• A confirmation and directions appear right after the booking is completed

What are common mistakes when reading conversion numbers?

The most common mistake is confusing "booking button clicks" with "completed bookings." Many customers tap the button and then just leave the booking screen, so the two numbers can differ quite a bit.

Also be careful not to watch only the conversion rate and miss the number of visitors. If visits dropped by half but the conversion rate rose, your actual bookings may have gone down.

Avoid overhauling your whole website the moment you have a bad month. A single month's numbers are easily swayed by chance, such as weather, closures or events, so look at the trend over at least two or three months before deciding.

• Double counting, where a conversion is recorded again every time the booking confirmation page is refreshed
• Comparing before and after while ignoring seasonal differences, like academy school breaks or a cafe's summer peak
• Reading a conversion rate change in a low-traffic month (e.g. from 2% to 4%) as a big win
• Setting so many key events that the numbers that matter get buried
• Not noting the dates you changed your website, so you can't tell what worked