Conversion rate is the share of visits or visitors that complete the action you care about, such as a purchase: conversions divided by the base, multiplied by 100. A store with 400 orders from 20,000 sessions has a 2% conversion rate on that basis. You do not need a conversion rate calculator for the division itself; the part that needs care is choosing what you divide by, because sessions, users and orders give different answers from the same month of data.
This guide covers the formula, worked examples with round hypothetical numbers you can redo on paper, how analytics and ad platforms define the metric, how to segment it, and what published benchmarks can and cannot tell you.
The formula
- Conversion rate (%) = conversions / base x 100
Two choices define the number:
- The conversion. For an online store it is usually a completed purchase. It can also be a sign-up, a quote request or an add to cart, but then it should be named as such (for example, add-to-cart rate) so nobody confuses it with sales.
- The base. Sessions (visits), users (people or devices), or ad interactions (clicks). Each answers a different question.
Always write the base next to the percentage. "2% conversion rate" means little until you know it is 2% of sessions.
How platforms define it
Google Analytics 4 offers two built-in versions. Session key event rate is the number of sessions in which a key event happened divided by the total number of sessions. User key event rate is the number of users who performed a key event divided by the total number of users. GA4 defines a session as a period during which a user interacts with your site or app, and by default a session times out after 30 minutes of inactivity.
Google Ads uses a different base: its conversion rate is conversions divided by the ad interactions (for example, clicks) that can be tracked to a conversion in the same period. Google's own example is 50 conversions from 1,000 interactions, a 5% conversion rate. Google also notes that this rate can exceed 100% when you count every conversion rather than one per interaction, because a single click can lead to several conversions.
So the same store can honestly report three different conversion rates on the same day. None is wrong; they measure different things.
Worked example: sessions vs users vs orders
All numbers here are hypothetical and rounded for easy arithmetic. In one month, a store records:
| Measure | Count |
|---|---|
| Sessions | 20,000 |
| Users | 14,000 |
| Orders | 400 |
| Sessions that contained at least one order | 390 |
| Users who ordered at least once | 380 |
Now three conversion rates from the same data:
| Rate | Calculation | Result |
|---|---|---|
| Orders per session | 400 / 20,000 x 100 | 2.00% |
| Session conversion rate (sessions with an order) | 390 / 20,000 x 100 | 1.95% |
| User conversion rate (users who ordered) | 380 / 14,000 x 100 | 2.71% |
Why they differ:
- Orders per session is slightly higher than session conversion rate because 10 sessions contained two orders each (400 orders in 390 sessions).
- User conversion rate is highest because many people visit more than once. 14,000 users made 20,000 sessions, so the base is smaller while the buyers are nearly the same.
The practical rule: pick one definition for reporting, write it down, and never compare a session-based rate from one tool with a user-based rate from another.
Segmenting conversion rate
A single store-wide rate hides most of what is useful. Break it down by segments you can act on:
- Device: mobile, desktop, tablet.
- Traffic source: paid search, paid social, organic search, email, direct, referral.
- New vs returning visitors.
- Landing page or product category.
- Country or language, which often tracks shipping cost and delivery time.
Worked example: device split
Taking the same hypothetical 20,000 sessions and 400 orders, split by device:
| Device | Sessions | Orders | Orders per session |
|---|---|---|---|
| Mobile | 14,000 | 196 | 1.4% |
| Desktop | 6,000 | 204 | 3.4% |
| Total | 20,000 | 400 | 2.0% |
Watch for mix shifts
Next month, suppose each device converts exactly as before, but traffic shifts toward mobile: 16,000 mobile sessions and 4,000 desktop sessions. Mobile produces 16,000 x 1.4% = 224 orders, desktop 4,000 x 3.4% = 136 orders. Total: 360 orders from 20,000 sessions, or 1.8%.
The store-wide rate fell from 2.0% to 1.8% even though neither segment got worse. The change came from the traffic mix, not the site. This is the most common reason a store-wide conversion rate moves, and the reason to check segments before redesigning a checkout.
Mind the sample size
Small segments swing wildly. With 100 sessions, 3 orders is 3% and 2 orders is 2%: a single order moves the rate by a full percentage point. Look at longer periods or combine small segments before drawing conclusions.
Benchmarks: use with care
Published averages are only useful if you know how they were measured. One primary source that states its method is Contentsquare's 2026 Digital Experience Benchmarks, published in March 2026. It draws on 99 billion web and app sessions across more than 6,000 sites and compares Q4 2024 with Q4 2025. In that data:
- Returning visitors converted at 2.9%, compared with 1.7% for new visitors.
- The desktop conversion rate was 74% higher than mobile, while mobile accounted for 69.9% of traffic.
- AI-referred traffic converted at 1.3%, up 55% year over year, just above organic social at 0.7%.
Keep three caveats in mind. First, the figures come from sites that use one vendor's analytics, across many industries, so they are not a typical figure for any single category. Second, they are measured on visits, so compare them only with your own session-based rate. Third, a benchmark tells you where others are, not what is achievable for your price point, product and traffic mix. Your own trend over time, by segment, is a far better yardstick.
Improving the number without fooling yourself
- Fix measurement first. Make sure every completed order is recorded once and only once, and that test orders are excluded.
- Pair conversion rate with revenue per session. Revenue divided by sessions captures both conversion rate and order value, so a discount that raises conversion but cuts order value shows up honestly. In the example, if the average order is $50, revenue is $20,000 and revenue per session is $1.00.
- Change one thing at a time and compare the same segment over the same length of period.
Where CIQRA fits
CIQRA shows revenue, orders, conversion and average order value on one live dashboard, and sends server-side conversions to GA4, Meta, TikTok, Google Ads, Pinterest, Snapchat and Microsoft. Whichever tool you report from, state the base you divide by.
FAQ
What is a good conversion rate for an online store?
It depends on your industry, price, device mix and traffic sources. Compare your rate with your own past performance by segment first, and treat published averages only as context.
Should I use sessions or users?
Sessions are the common choice for day-to-day reporting and match most published benchmarks. Users are better when you care whether a person eventually bought across several visits. Pick one and label it.
Why does my ad platform show a different conversion rate than my analytics?
They usually use different bases (ad interactions vs sessions), different attribution and different counting rules, so the numbers are not expected to match.
Sources
- Google Analytics Help: [GA4] Dimensions and metrics (accessed 2026)
- Google Analytics Help: About Analytics sessions (accessed 2026)
- Google Ads Help: Conversion rate: Definition (accessed 2026)
- Contentsquare: Conversion benchmarks: tracking growth in a high-cost, high-intent world (2026 Digital Experience Benchmarks) (2026)