The Complete Guide to Benchmarks

Most teams use benchmarks badly. They pull one aggregate number from a blog post, compare it to their own dashboard, and either celebrate or panic — without checking whether the segments match. A median ecommerce conversion rate of 2.5% includes everything from a $9 phone case to a $4,000 mattress, and those funnels behave nothing alike.

The benchmarks worth tracking sit one layer down. What's a normal mobile checkout completion rate for a beauty brand with a €45 AOV? What ROAS should a Meta-led apparel store expect in month four after iOS attribution loss? Those are the questions this hub answers, with the underlying spokes — Ecommerce Conversion Benchmarks, Mobile Conversion Benchmarks, Checkout Benchmarks, ROAS Benchmarks, CAC Benchmarks — going deeper on each cut.

Treat the numbers below as orientation, not targets. They tell you whether something deserves a test, an audit, or no action at all. Your store's own 90-day trend matters more than any external median — but you need the external median to know what's worth chasing in the first place.

Ecommerce conversion benchmarks

Site-wide conversion rate is the headline number every operator watches, and the one most often misread. The global median sits around 2.0-2.5% for online retail, but the spread by vertical is enormous: consumer electronics and food often crack 3-4%, while furniture, jewellery and high-AOV apparel sit closer to 0.8-1.5%.

What moves the number isn't usually the headline figure — it's the segment underneath. Top-quartile beauty stores convert paid traffic at 2.5-3.5% and email traffic at 6-9%; their site-wide average looks unremarkable because cold prospecting drags it down. Our Ecommerce Conversion Benchmarks and Fashion Ecommerce Benchmarks spokes break this apart by traffic source and AOV tier.

The other variable people forget is intent. Returning visitors convert 3-5x higher than first-touch traffic on most stores. If your returning-visitor share dropped from 35% to 22% after a paid push, your site-wide rate will fall even if every page got better. Always cut conversion by new vs returning before drawing conclusions.

Mobile vs desktop: the gap that won't close

Mobile drives 65-75% of ecommerce sessions but only 45-55% of revenue on most stores. The conversion-rate gap between mobile and desktop has narrowed slightly since 2020 but hasn't closed — desktop still converts roughly 1.6-2.0x higher across most verticals.

The gap isn't a screen-size problem; it's a context problem. Mobile sessions are shorter, more interrupted, and more research-led. Desktop sessions skew toward checkout-ready intent. That means the right benchmark target isn't 'close the gap' — it's 'protect mobile add-to-cart while moving the purchase to whichever device the shopper prefers'. Our Mobile Conversion Benchmarks spoke covers device-specific stage rates in depth.

Where mobile genuinely underperforms is checkout. Mobile cart abandonment runs 5-10 points higher than desktop, and form completion times are 30-50% longer. Those are the gaps worth attacking with Apple Pay, Shop Pay, address autocomplete and one-page checkout.

Checkout and cart abandonment

Cart abandonment is the single most-quoted benchmark in ecommerce and one of the most misleading. The famous '70% of carts are abandoned' figure averages across hard intent (shoppers who entered payment details) and casual intent (shoppers who used cart as a wishlist). Treat the funnel stage by stage instead.

Realistic ranges: 60-75% of carts never reach checkout, then 25-40% of checkout-starts never complete. The second number is the one that signals UX problems — high cart-page exits usually mean shipping cost surprises or account-creation friction, while high checkout-form drop-off usually means payment, validation or trust gaps. The Checkout Benchmarks and Cart Abandonment Benchmarks spokes split these out by step.

Vertical matters too. Beauty and apparel sit near the global average; furniture, jewellery and considered electronics run 5-15 points higher because of price and research cycles. The table below gives baseline conversion rates across verticals — use them as the floor for what 'normal' looks like before benchmarking your own funnel.

If your numbers sit between median and top-quartile, the next gain is usually segment-specific — not a site-wide redesign. Find the one device, source or product line dragging the average and fix that. Our Beauty Ecommerce Benchmarks and Fashion Ecommerce Benchmarks spokes show what the top-quartile teams do differently.

Channel acquisition: CAC and ROAS

Acquisition benchmarks have shifted hard since 2021. Meta CAC for DTC stores roughly doubled from 2020 to 2024; blended ROAS targets that used to be 4-5x are now 2.2-3.0x for sustainable growth, with first-order ROAS often dipping below 1.5x and the rest of the unit economics carried by repeat purchase.

Healthy channel ranges depend on margin and repeat rate, but as orientation: paid social ROAS of 1.8-3.0x, paid search of 3-6x (branded skews much higher), email of 25-50x, and organic/SEO functionally infinite once amortised. CAC payback under 4 months is strong for repeat-heavy verticals like beauty; under 8 months is normal for considered purchases. The ROAS Benchmarks and CAC Benchmarks spokes go deeper, and Funnel Benchmarks ties channel performance to on-site stage rates.

The mistake here is comparing your channel ROAS against a universal number instead of your own contribution-margin floor. A 2.5x ROAS is great at 65% margin and terrible at 30%. Always benchmark CAC against LTV and margin together, not in isolation — Conversion Rate Optimization and Ecommerce Metrics cover this end-to-end.

Experimentation velocity and SaaS benchmarks

How fast you can learn matters more than how good any single test is. Top-quartile experimentation programmes ship 4-8 meaningful tests per month, win 18-25% of them, and lift the metric they target by 4-9% on average. Most stores ship fewer than one test a month and call themselves a CRO operation.

Sample size is the constraint. A store doing 50k monthly sessions and 2% conversion needs roughly 3-5 weeks per test for a 10% relative lift at 80% power. That math is unforgiving — and it's why test prioritisation and hypothesis quality outrank raw test volume. The Experimentation Benchmarks spoke covers velocity, win rate and required traffic by store size.

Adjacent reference points: SaaS Conversion Benchmarks and SaaS Activation Benchmarks for software businesses, Industry Benchmarks for cross-vertical comparison, and the metric definitions in Ecommerce Metrics if you need to ground the vocabulary first. Every linked spoke uses the same segmentation logic — vertical, device, channel, stage — so the cuts compose.