How to Calculate Website Conversion Rate
Calculate website conversion rate with matching users or sessions, clear conversion definitions and worked examples. Avoid duplicate events and mixed scopes.
To calculate website conversion rate, divide the number of users or sessions that completed your chosen action by the matching eligible users or sessions, then multiply by 100. Decide what counts as a conversion and which denominator you are using before doing the division.
For example, 60 sessions containing a purchase out of 3,000 eligible sessions gives a purchase session conversion rate of 2%.
On a WordPress website using Niwa, the useful action might be a purchase after product assistance or a confirmed website booking after a service question. Opening the assistant is a separate interaction. Keeping those outcomes separate makes the calculation meaningful.
1. Define the action you are counting
Write a short measurement definition, such as: "A completed purchase recorded for a valid order during the reporting period." For a service business, use a saved reservation rather than a click on the booking button.
If you measure enquiries, decide whether you are counting all successfully received submissions or only qualified enquiries. Both can be useful, but they answer different questions.
A purchase event should correspond to the actual completed transaction. Repeated page loads and duplicate tracking should not create new orders. For bookings, test that the appointment really exists in the schedule before counting the event as success.
2. Choose users or sessions
For a session-based rate, count sessions containing at least one instance of the conversion:
Converting sessions ÷ eligible sessions × 100.
For a user-based rate, count distinct measured users who converted:
Converting users ÷ eligible measured users × 100.
These denominators are not interchangeable. In a hypothetical period, 1,000 measured users make 1,500 sessions. If 30 users convert in 35 sessions, the user rate is 3%, while the session rate is approximately 2.33%.
Five extra converting sessions do not imply five extra customers. Some people returned and converted again.
If you divide the number of orders by sessions, label the result orders per session. It can be useful, but it is not necessarily the percentage of sessions that converted. A session can contain more than one order.
3. Keep the numerator and denominator in the same scope
Use the same dates, timezone, website, device or campaign segment for both parts of the fraction. If the numerator is mobile purchases, the denominator should describe the mobile population you are evaluating.
For a checkout completion rate, the denominator might be sessions that started checkout. For a whole-store purchase conversion rate, it could be all eligible store sessions. Do not compare those rates as though they measure the same journey.
Define your reporting exclusions consistently. Test traffic and known duplicate events should be investigated. Do not selectively remove ordinary non-buyers simply to increase the percentage.
4. Read the correct rate in GA4
Google Analytics calls important measured actions key events. Its Traffic acquisition report documentation defines session key event rate using sessions where a key event happened, rather than the total number of times it happened.
Choose the specific event you need, such as purchase, when your report offers an event selector. An "any key event" rate can combine several actions, including intermediate ones you have marked as important. It should not be relabelled as your purchase rate.
Configure and test the required event, then mark it appropriately in your report. Note changes in tracking or event definitions so the rates remain comparable over time.
5. Compare percentages correctly
Suppose the conversion rate rises from 2% to 2.4%.
- The absolute change is 0.4 percentage points.
- The relative increase is 20%, calculated as (2.4 − 2) ÷ 2 × 100.
Always include the underlying counts. A rate of 10% based on one converting session out of ten is much less informative than 10% based on 1,000 out of 10,000. Neither example by itself proves an improvement caused by a website change.
If the denominator is zero, the rate is undefined. Show "no eligible traffic" rather than reporting a fabricated zero-performance result.
Measure the outcome after assistance
Niwa helps shoppers compare products and complete sales workflows using the business information on your website. Record the completed order separately from the question, recommendation and cart action. For services, Niwa can answer questions, check availability and complete a reservation directly in WhatsApp. Count the saved booking as the outcome. In a website conversion rate, include only outcomes connected to the eligible website population under your measurement rules; report bookings completed entirely in WhatsApp separately.
You may also report conversion rates for assisted and unassisted visitors, but those groups choose their own behaviour. A higher assisted rate is an association, not automatic evidence that the assistant caused the difference.
Use Niwa's storefront assistance or booking workflow to support a defined customer task, then measure that task with the same rules over time. Send Niwa your current website and target outcome if you need help choosing a suitable WordPress setup.
