Definition
Conversion rate
The share of sessions or visitors that complete a defined action, usually a purchase, in a given period.
Conversion rate is the share of people who complete a defined action out of everyone who had a chance to. In ecommerce the default action is purchase: orders divided by sessions, or buyers divided by visitors, depending on how you count. It is the bridge between traffic cost and revenue.
Raise conversion on the same ads and effective customer acquisition cost falls; leave conversion flat and every CPM hike hits the bank account. It is not a moral score for your brand, and it is not a single “industry average” you must match. It is a ratio with a numerator, a denominator, a device mix, and a sample size. Misread any of those and CRO work turns into noise.
The purchase conversion formula
The standard storefront formulas are: **Session conversion rate = number of orders ÷ number of sessions** **Visitor conversion rate = number of converting customers ÷ number of unique visitors** in the same period. Shopify’s analytics commonly lean on sessions; other tools emphasize users. Do not mix them in one trend line. An order is a completed purchase (define whether you exclude fraud cancels or 100% refunds on day zero).
A session is a visit bundle under your analytics rules; a visitor is a deduplicated person or browser under those same rules. Example: 120 orders and 10,000 sessions → 1.2% session conversion. That percentage only moves when orders or sessions move. A viral spike of low-intent sessions can drop conversion while revenue rises. Both can be true.
Always pair conversion with revenue, average order value, and order count so you do not “win” a rate while losing the business.
Macro vs micro conversions
Macro conversion is the outcome that pays the bills, almost always purchase, sometimes a qualified lead for high-consideration goods. Micro conversions are intermediate actions: add to cart, begin checkout, email signup, quiz complete, or subscription attach. Micro rates diagnose where the funnel breaks; they are not a substitute for purchase conversion in executive reporting.
A healthy practice is a short funnel table: session → product view → add to cart → checkout started → purchase. Each step has its own conversion rate from the previous step. When add-to-cart is fine but checkout start is weak, the problem is often cart UX, shipping visibility, or trust, not PDP creative. When checkout start is fine but purchase fails, payment, validation, or performance is the suspect.
Abandoned cart recovery sits after the miss; funnel conversion work tries to prevent the miss. Report both, and do not let a signup-rate win hide a purchase-rate loss.
Mobile, desktop, and channel mix move the number
Storewide conversion is a weighted blend. Mobile traffic often converts lower than desktop for many catalogs, not because mobile is “bad,” but because intent, screen constraints, and checkout friction differ. If your session mix shifts toward mobile after a social campaign, blended conversion can fall even when mobile conversion is stable. Paid prospecting usually converts lower than branded search or email; a budget shift changes the denominator quality overnight.
Segment before you panic. Break purchase conversion by device, new vs returning, and primary channel. Returning and email traffic will flatter the average; cold paid will depress it. CRO tests should specify which segment they target. Improving desktop checkout does little if 80% of sessions are mobile. Improving paid landing pages does little if the chart you watch is dominated by email.
The metric is honest only when the mix is visible next to it.
Sample size, seasonality, and false wins
Conversion rate is a proportion estimate. Small catalogs and low-traffic days produce noisy percentages. A jump from 1.1% to 1.4% on 800 sessions is often random; the same jump on 80,000 sessions is more interesting. Run tests long enough to cover weekday and weekend patterns, and avoid calling winners in the middle of a one-day flash sale unless that sale is the permanent reality. Seasonality and inventory also distort the ratio.
Stockouts lower conversion without a UX regression. Shipping cutoffs before holidays change behavior. Coupon codes spike conversion and can train discount dependence. I require a pre-registered primary metric, a minimum runtime or sample rule, and a check that AOV and margin did not collapse. A conversion lift funded by a 30% sitewide code is not free efficiency; it is a different offer.
Document the offer context next to every “record conversion week” screenshot.
CRO levers that actually move purchase rate
Conversion rate optimization is structured change to pages, offers, and flows, measured against purchase or a declared micro metric. Highest-impact areas for most stores: PDP clarity (price, variants, shipping promise, social proof that matches the SKU), site speed, search and collection merchandising, cart and checkout friction, payment options, and trust cues that reduce uncertainty, not generic stock urgency timers. Cross-metric effects matter.
Bundles and thresholds can raise average order value while slightly lowering conversion; that trade can still be correct on contribution. Free-shipping thresholds can lift conversion for cart values near the line and stall smaller carts. Cutting steps in checkout often helps mobile more than desktop. Fix abandoned cart causes upstream (surprise shipping cost, forced account creation, slow payment) before you scale recovery emails.
Recovery is a net; conversion rate is whether the cart was catchable at all.
How conversion rate ties to CAC and growth math
Traffic teams buy sessions or clicks; the store turns them into customers. Effective cost per acquisition moves roughly with media cost per session divided by conversion rate (and with how you define customers). If CPM rises and conversion stays flat, customer acquisition cost worsens. If conversion rises on the same traffic quality, you buy more customers per dollar without negotiating a single auction.
That is why CRO and media are one system inside the ecommerce growth stack, not rival departments. Landing-page relevance is both a conversion problem and a CAC problem. So is mobile checkout. So is inventory accuracy on the PDP. When reporting weekly, put session conversion, new-customer share, AOV, contribution, and blended CAC on the same page. A conversion win that only appears on returning email traffic will not rescue paid prospecting CAC.
Optimize the path that matches the money you are trying to spend.
Common questions
Frequently asked questions
How do you calculate ecommerce conversion rate?
Divide completed purchases (or another defined action) by sessions or unique visitors in the same period, then multiply by 100 for a percentage. Pick sessions or visitors deliberately and do not switch mid-trend.
What is a good ecommerce conversion rate?
There is no single good rate for every store. Compare against your own baseline by device and channel, and against peers only when price point, traffic quality, and category are similar, not against a generic global average treated as truth.
What is the difference between macro and micro conversion?
Macro conversion is the primary business outcome, usually purchase. Micro conversions are intermediate steps such as add to cart or checkout started. Use micro rates to diagnose the funnel; use macro rates to judge the business.
Why did conversion rate drop when traffic went up?
Often traffic quality or mix changed (more cold paid or mobile sessions) while orders did not rise proportionally. Stockouts, site issues, or offer changes can also drop the rate. Segment by channel and device before redesigning the whole site.
How does conversion rate affect CAC?
For a given cost per session, higher purchase conversion produces more customers per dollar and lowers effective CAC. Conversion work on the paths you actually buy traffic for is one of the cleanest ways to improve acquisition math.
Related terms
