Plain-English definitions of the terms we use when we talk about conversion rate optimization, Shopify, and DTC ecommerce.
A controlled experiment comparing two variants of a page to determine which performs better.
A/B testing randomly splits traffic between a control and one or more variants, then measures the difference in a target metric (usually conversion rate or revenue per visitor). It's the most rigorous way to validate a change, but it requires sufficient traffic, statistical discipline, and a clear hypothesis.
The percentage of sessions that result in a product being added to the cart.
ATC rate is the first meaningful intent signal in the funnel. If ATC rate is high but conversion rate is low, the problem is likely in cart or checkout. If ATC rate is low, the problem is upstream - product page, pricing, or trust signals.
The average amount a customer spends per order. Total revenue ÷ total orders.
AOV is one of the three levers of ecommerce revenue (traffic × conversion rate × AOV). Raising it is often faster than raising conversion rate: upsells, bundles, free-shipping thresholds, quantity discounts, and post-purchase offers all move AOV without acquiring a single new customer.
Example
If 100 orders generate $7,500 in revenue, AOV is $75.
The percentage of sessions where a user leaves without any interaction.
High bounce rate isn't always bad - some pages (like blog articles) answer the user's question in one visit. On product and landing pages, though, a high bounce rate usually signals a mismatch between ad creative and page content, or friction in the above-the-fold experience.
When a shopper adds to cart but never starts checkout.
Cart abandonment is usually a commitment problem, not a checkout problem. The fix is upstream: clearer pricing, visible totals, shipping cost transparency, and a persistent cart indicator.
When a shopper reaches checkout but leaves before completing the order.
Industry average checkout abandonment is 60-70%. Common causes: unexpected shipping costs, forced account creation, limited payment options, or slow checkout load times. Recovering even a fraction via better UX + abandoned-cart email flows produces significant revenue.
The sequence of steps a visitor takes from landing on the site to completing a purchase.
A typical Shopify funnel: landing page → collection page → product page → cart → checkout → thank-you. Every step drops a percentage of users. CRO finds the biggest drop and fixes it first - usually with more impact than optimizing a step that's already working well.
The practice of systematically increasing the percentage of site visitors who complete a desired action - usually a purchase.
CRO is a structured process of research, hypothesis, experimentation, and iteration aimed at removing friction and amplifying motivation across a site's funnel. For Shopify DTC brands, CRO typically focuses on product pages, collection pages, cart, and checkout - the stages where most revenue is won or lost. Done properly, CRO compounds: every winning test stacks on the previous one, so gains accumulate over months.
Example
A 2% → 2.4% conversion rate on $1M/mo in traffic is +$200k/yr in revenue.
The total cost to acquire one paying customer, including ad spend, agency fees, and tooling.
CAC is measured against LTV - a healthy LTV:CAC ratio for DTC is typically 3:1 or higher. When CAC rises (as it has across Meta and Google since 2022), brands that can't lift conversion rate or AOV quickly go from profitable to unprofitable.
The total revenue a single customer is expected to generate across their entire relationship with a brand.
LTV determines how much you can afford to spend on acquisition (CAC) while remaining profitable. Brands with strong LTV - through subscriptions, repeat purchase, or high AOV - can outbid competitors on ads and still grow profitably. Improving LTV is usually cheaper than improving CAC.
Example
A customer who places 4 orders of $75 over 18 months has an LTV of $300.
A visual representation of where users click, tap, or move on a page.
Heatmaps show attention and engagement, not intent. They're useful for spotting obvious issues (dead zones, ignored CTAs, confused clicks on non-clickable elements) but shouldn't be used alone to make design decisions.
A falsifiable prediction about how a specific change will affect a specific metric and why.
A strong CRO hypothesis has three parts: the change, the expected effect on a metric, and the reason. 'Because users on mobile lose the CTA on scroll, adding a sticky Add to Cart bar will increase mobile ATC rate.' Without a reason, a test is guessing.
The percentage of mobile sessions that convert, usually significantly lower than desktop.
Mobile traffic is typically 60-80% of a DTC store, but mobile converts at roughly half the rate of desktop. Closing that gap is one of the biggest revenue levers available. Sticky CTAs, faster load times, simplified variant selectors, and better checkout UX move the number most.
The page for a single product, containing images, description, variants, pricing, and the Add to Cart button.
PDPs are where most ecommerce revenue is won or lost. The above-the-fold area needs to answer three questions on load: what is it, why should I buy it, and how much does it cost? Everything else is secondary.
A page showing multiple products, usually a collection, category, or search result.
PLPs serve differentiation - helping shoppers identify the right product quickly. Badges (bestseller, best value), spec callouts, and strong product imagery are what separate a PLP that sells from one that overwhelms.
Total revenue divided by total unique visitors. The single cleanest measure of funnel health.
RPV combines conversion rate and AOV into one number, so it's harder to game than either metric alone. When you run a CRO test, RPV is the metric that matters - a higher conversion rate at a lower AOV can actually reduce revenue.
The number of sessions or conversions required for a test to produce a reliable result.
Sample size depends on baseline conversion rate and the lift you want to detect. A store at 2% CR needs far more traffic to detect a 5% relative lift than to detect a 30% lift. Underpowered tests produce noise, not signal.
A video replay of a real user's visit to the site.
Session recordings reveal why users behave the way they do - rage-clicks, dead-end scrolls, form friction, and confusion that analytics alone can't explain. Used together with quant data, they're one of the highest-signal research tools in CRO.
Shopify's enterprise tier, designed for brands doing roughly $1M+/yr in revenue.
Shopify Plus adds checkout customization, multiple stores under one account, higher API limits, scripts, launchpad, dedicated support, and B2B features. It's worth upgrading when you hit specific constraints - most often checkout customization or store scale - not just because revenue is high.
A measure of confidence that an A/B test result is not due to random chance.
The industry convention is 95% confidence. Calling a test early - before it reaches significance - is the single most common CRO mistake and produces false winners that hurt the business. Significance also requires sufficient sample size: small traffic stores often can't run traditional A/B tests and need different test structures (sequential testing, Bayesian approaches, or before/after holdouts).
Example
A 20% lift with 87% confidence is not significant. Either continue the test or move on - never ship it.
Ready to apply these?
We turn these concepts into revenue for Shopify DTC brands every month - through ongoing CRO and high-converting landing pages.
