Conversion Rate Formula and What One Extra Point Is Worth

Conversion rate is conversions divided by the chances to convert, times 100. The chances differ by system: Google Ads divides by ad clicks, GA4 by sessions, your shop by visits. Judge ads and bids by the Google Ads rate, the whole site by the shop’s, and find drop-off points in GA4. One extra percentage point raises ROAS by 1 ÷ your current rate. From 2.7% to 3.7%, ROAS rises 37% at the same average order value and cost per click, or you keep ROAS and pay 37% more per click.

Your conversion rate joins the ads and the site: the ads bring the click, the site makes the sale. The guide to conversion rate optimisation for online stores shows how to raise it. This article shows how to measure it and what one extra point is worth.

What is the conversion rate formula?

Conversion rate = conversions ÷ chances to convert × 100%

In an ad account the chances are ad clicks, so the rate is the share of clicks that ended in an order. On the site they are visits, so the rate is the share of visits that ended in an order. The formula stays the same. What you put above and below the line changes.

Example store, not client data.

A tableware shop with 3,000 products gets 8,000 clicks from Google Ads in a month, and 300 of them end in an order. Its conversion rate from ads is 300 ÷ 8,000 × 100% = 3.75%.

Two rules keep the number honest:

• Numerator and denominator come from one system, one period and one slice of traffic. Orders from every channel divided by clicks from one channel is not a rate.
• Conversions are orders, not steps towards them. An add to cart or a checkout start in the numerator measures interest, not sales.

Why do Google Ads, GA4 and your shop show different conversion rates?

Because each system divides different things. Here is how each one defines its rate.

Google Ads. The Google Ads API metrics reference defines Conv. rate as conversions from interactions divided by the number of ad interactions. An interaction is the main action for the ad format: a click for text and Shopping ads, a view for video ads. The numerator holds only the actions included in the Conversions column, and those are what conversion-based bid strategies optimise for.

If a campaign also serves video ads, their views enter the denominator. For product ads, conversions ÷ clicks is the cleaner figure, and it is how we count our store benchmarks.

GA4. The closest metric is session key event rate: the percentage of sessions in which any key event fired, according to Google’s API dimensions and metrics list. The word “any” matters. Purchase is a key event by default, but if you also mark an add to cart or a newsletter sign-up as a key event, those sessions count too.

The same list has a version for a single key event: set it to purchase, and it counts only sessions with a purchase. Purchaser rate is different again: the share of active users who made at least one purchase. Its denominator is people, not sessions.

Your shop. Shopify’s guide (Ecommerce Conversion Rate: Benchmarks & Tips) defines the rate as orders divided by visits to the site. It adds that you should count visitors as sessions, not users. The shop sees every order, whichever channel it came from: ads, organic search, email or a direct visit.

Shopify’s own report counts a little differently. Its Help Center (Behavior reports) defines the rate as the percentage of sessions that resulted in a purchase. Two orders in one session therefore count once.

Example store, not client data.

In the same month the tableware shop records 21,000 sessions from all channels and 546 orders. In GA4, filtered to Google Ads traffic, it has 7,500 sessions, 270 of them with a purchase.

Each of the first three rates is right for its own question. The fourth is a common error: orders from every channel divided by clicks from one. The article on Google Ads vs GA4 conversions explains why the two Google systems can count the purchases themselves differently.

Which conversion rate should you use?

Pick the rate by the decision it feeds.

Whichever you use, compare it first with your own previous periods of the same length. A rate that moves within one system is a signal. A gap between two systems may be nothing more than a difference in definitions.

When the formula is right but the number is wrong

The division is rarely the problem. The numerator is. These four things change the conversion count without a single extra order:

1. A funnel step counted as a conversion. If an add to cart sits next to the purchase in the Conversions column, the rate counts carts too. GA4’s session key event rate does the same when you mark extra key events.
2. Two actions counting one order. An old tag and an imported goal that both feed the Conversions column double the numerator.
3. The counting setting. A Google Ads conversion action counts either all conversions per click or only one per click (ConversionActionCountingType). With one per click, Google Ads leaves out a second order after the same click, so the ads show fewer orders than the shop.
4. No transaction ID in GA4. Google’s event reference says the transaction_id parameter of the purchase event helps you avoid duplicate purchase events (Recommended events). Without it, GA4 can count a repeated purchase event as a new order.

The portal’s conversion check covers part of this list. It checks whether the ads learn on a purchase rather than a page view or an add to cart, and whether real order amounts arrive. It also looks for two goals counting the same order and for single orders many times your average order value.

Check the counting setting yourself. The portal reports what it finds and changes nothing in your account.

Half of the stores in our data convert 1.9–4.0% of ad clicks

We took GetProfit data on 114 stores with at least eight months of history, June 2025 – June 2026. For each store, we divided conversions by clicks in Shopping and Performance Max product ads over the whole window. The lower quartile is 1.9%, the median 2.7% and the upper quartile 4.0%. Half of the stores sit between the two quartiles.

Our study of 1.4 million products, 130+ stores over 13 months, puts the median conversion rate at 2.29% and the average at 3.65%. The samples and windows differ, and so do the two medians.

The gap between the median and the average in the study is the more useful lesson: a few stores with very high rates pull the average up. Compare your store with the median, not the average. Compare it per ad click, too: that is how we counted the 114-store figures, and the study also uses ad account data rather than shop analytics.

The article on average ecommerce conversion rate covers what rate is normal and whether a higher one means a more profitable store. One finding from the same 114 stores matters here, because it changes how you value one extra point.

Across stores, a higher conversion rate does not mean a higher ROAS. Split the stores into quarters by ROAS, and the top quarter (median ROAS 1,228%) converted a median 2.75% of clicks, less than the second quarter’s 3.13%.

Within one store, the picture changes. Across 1,360 store-months, a store’s ROAS moved with its conversion rate (rank correlation +0.527) and its average order value (+0.540). It barely moved with its cost per click (−0.100).

In the 361 months when a store’s ROAS rose more than 20% against other stores that month, conversion rate rose a median 23.3% and average order value 19.5%. Both rises are also measured against other stores. These are observations, not an experiment: they show what moves together, not what causes what.

So to see what one point is worth, compare your store with itself.

What is one extra point of conversion rate worth?

Conversion rate sits inside ROAS. Revenue is clicks × conversion rate × average order value. Spend is clicks × cost per click. The clicks cancel out:

ROAS = conversion rate × average order value ÷ cost per click

Example store, not client data.

For the tableware shop, 3.75% × 600 ÷ 5 = 4.5. That is the same ROAS as 180,000 in revenue ÷ 40,000 in spend. Now add one point of conversion rate and keep everything else the same.

One point brings 80 more orders for the same money. Spend stays where it was, so the shop keeps the whole 16,800 of extra gross profit. What is left after ad spend grows by 73%.

The shop can also spend that point instead of keeping it. To hold ROAS at 4.5, it can now pay 4.75% × 600 ÷ 4.5 = 6.33 per click instead of 5, which is 26.7% more room to bid. The ceiling where ads eat the whole margin rises the same way: from 7.88 to 9.98 per click. The article on break-even CPC works out that limit for your own margin.

The lower your rate, the more one point is worth

The gain from one extra point is 1 ÷ your current rate. Applied to the quartiles of the 114 stores:

Percentage points and per cent are two different things. One percentage point takes 2.7% to 3.7%. One per cent takes 2.7% to 2.727%, which adds just 1% to ROAS. When someone promises “+1% conversion rate”, ask which of the two they mean.

When the arithmetic stops holding

The ROAS formula assumes the same clicks, the same average order value and the same cost per click. In practice one of them often moves with conversion rate:

• A discount can lift conversion rate and lower average order value. ROAS then rises less than the table shows, or not at all.
• A shift in traffic changes what the clicks cost. If you raise bids on the products that convert best, cost per click can rise along with the rate.
• A change in counting lifts the rate with no new orders. If a second conversion action starts counting the same purchases, only the report changes.

So track conversion rate, average order value and cost per click together, from the same period.

How to calculate your store’s conversion rate, step by step

1. Pick whole weeks and leave out the last few days. Google’s Conversion reporting guide notes that performance data isn’t available instantly, so the latest days are incomplete. Compare windows of equal length.
2. In Google Ads, take Conversions and Clicks for your Shopping and Performance Max campaigns, then divide Conversions by Clicks. Check that only the purchase action feeds the Conversions column. If you need conversions dated by the day they happened, Google Ads has separate “(by conv. time)” versions of the columns.
3. In GA4, use the rate for the purchase key event, not for all key events, and filter to the traffic you want to compare. GA4 counts a purchase only when the site sends a purchase event, as set out in Google’s Measure ecommerce guide.
4. In your shop, divide orders by sessions for the same dates.
5. Write the three rates side by side with their denominators, and track each one against its own history.
6. Run your numbers through the ROAS formula. Take conversions ÷ clicks, value per conversion and average cost per click from the same Google Ads view. Multiply the first two, divide by the third and check that the result matches your ROAS. Then add one point and see how ROAS moves.
7. Before you compare anything, check the conversions themselves: one purchase action, real order amounts, no duplicate goals.

See what share of your numbers you can trust. Sign in with Google in one click. The portal changes nothing without your consent.

Check your conversions →

Sources

• metrics | Google Ads API — Conv. rate is conversions from interactions divided by ad interactions; an interaction is a click for text and Shopping ads and a view for video ads; conversions include only actions in the Conversions column, which bid strategies optimise for; value per conversion and average CPC definitions. Checked 2 October 2026.
• Conversion reporting | Google Ads API — the Conv. rate column maps to conversions_from_interactions_rate; the “(by conv. time)” columns; performance data isn’t available instantly. Checked 2 October 2026.
• ConversionActionCountingType | Google Ads API — count all conversions per click or only one per click. Checked 2 October 2026.
• API dimensions and metrics | Google Analytics — session key event rate, its single-event version, sessions, purchaser rate, cart-to-view and purchase-to-view rates; purchase is a key event by default. Checked 2 October 2026.
• Recommended events | Google Analytics — transaction_id in the purchase event helps avoid duplicate purchase events. Checked 2 October 2026.
• Measure ecommerce | Google Analytics — a purchase is measured by sending a purchase event. Checked 2 October 2026.
• Ecommerce Conversion Rate: Benchmarks & Tips (2026) — the shop formula, orders divided by visits; visitors counted as sessions, not users; the “online store conversion rate” name in Shopify Analytics. Checked 2 October 2026.
• Behavior reports | Shopify Help Center — Shopify’s conversion rate is the percentage of sessions that resulted in a purchase; order count and sessions that completed checkout can differ. Checked 2 October 2026.
• GetProfit data: 114 stores with at least eight months of history, June 2025 – June 2026 — conversion rate per ad click by quartile, conversion rate by ROAS quarter; ROAS = conversion rate × average order value ÷ cost per click.
• GetProfit data: 1,360 month-to-month changes in store ROAS, June 2025 – June 2026, each measured against the median store that month — what moves with ROAS within a store.
• GetProfit study of 1,404,808 products, 130+ stores, 13 months — median conversion rate 2.29%, mean 3.65%.
• GetProfit portal methodology — what the conversion check looks at.