The conversion rate of Selin's home textiles store is 1.1%. When she read "the average is 3%" in a marketing newsletter, she spent two days convinced her store was broken. Then another article said "1-2% is normal" and she relaxed. Both numbers were quoted from somewhere; neither said which stores they were measured on, over which period, or by which definition.
I invented Selin for this article; I did not invent her question — "is my rate any good?" is the question we hear most often in e-commerce meetings. Instead of making up a number to answer it, we opened public datasets with published methods one by one: nine sources, each with its date, measurement period and sample. None of the figures in this article come from INDOLES clients; we do not have a shareable, auditable client dataset and we do not pretend to. Our contribution is compilation and interpretation: which number means something for whom, and which one misleads.
The figures were read from their sources on 25 September 2026. Most sources publish rolling data, so some may have changed by the time you look. Each source's details are at the end of the article.
What is the average e-commerce conversion rate?
Across large public datasets the average e-commerce conversion rate sits between 1.4% and 2.72%; Dynamic Yield, the broadest and most recent global series, gives 2.72% for August 2025 to July 2026. Roughly, two or three in every hundred visits end in an order and the rest do not.
- Dynamic Yield by Mastercard: 2.72% — a 12-month global average across the company's customer base of 200 million monthly unique users and 300 million sessions.
- IRP Commerce: 2.23% — August 2026, independent SME and mid-market stores in the UK and Ireland; 1.85% in the same month a year earlier.
- Triple Whale: 1.69% — August 2025 to July 2026, the median conversion rate of paid-ad traffic across more than 53,000 brands.
- Littledata: 1.4% — the average of 2,800 Shopify stores measured in 2023; the top 20% sit above 3.2% and the top 10% above 4.7%.
- Contentsquare: 3.4% on desktop, 2% on mobile — the last quarter of 2025, more than 6,500 sites in 9 sectors and 99 billion sessions; it does not publish a single overall average.
None of these five numbers is wrong, but none of them is your store's number either. The gap between 1.4% and 2.72% comes less from stores being good or bad than from who counts what, divided by which denominator. The good news: once you know where the gap comes from, you also know which number sits closest to you — which one is the right benchmark for you.
Why do the sources give different averages?
Because each source looks at a different denominator, a different definition of conversion and a different sample of stores. The same store's rate can double depending on the definition used; comparing two numbers whose definitions do not match produces noise, not information.
Denominator: sessions or visitors?
IRP Commerce calculates the rate by dividing transactions by sessions. Dynamic Yield divides completed purchases by visitors. The difference looks small but is not: someone who visits three times in a week and buys once scores 33% on a session basis and 100% on a visitor basis. In a store with many returning visitors, the visitor-based rate comes out clearly higher than the session-based one.
Numerator: purchases only, or any conversion?
Contentsquare defines conversion as "the share of visits that include a conversion event, like a purchase or form submission"; newsletter sign-ups and forms can count. Dynamic Yield, IRP and Littledata count purchases. Triple Whale's 1.69% is the rate of traffic from paid-ad channels only, not the whole site.
Sample: who produces the average?
Dynamic Yield's data comes from its own customers; brands that have budgeted for personalisation software are, on our reading, retailers above a certain scale. Littledata measures Shopify stores, IRP independent SMEs in the UK and Ireland, Triple Whale brands that have connected their ad data. Contentsquare's 9 sectors include non-retail areas such as finance and travel. Even the same sector name can mean different stores: food and beverage is 4.80% at Dynamic Yield and 1.58% at IRP.
Period and statistic: mean or median?
Dynamic Yield gives a 12-month average, IRP a single month, and Contentsquare compares the final quarters of two years. Triple Whale uses the median rather than the mean; the median stops a few very successful brands from pulling the figure up and sits closer to the typical store. Comparing a period that includes November with one that does not can by itself create a gap of more than half a point; you will see the figure below, in the section on comparing your own rate.
How much does conversion rate vary by sector?
Sector is the single variable that moves conversion rate most: in Dynamic Yield's 12-month data, beauty and personal care leads at 5.39% and luxury and jewellery trails at 0.72%, a gap of more than sevenfold. Frequently bought, low-value products convert high; rarely bought, long-considered products convert low.
- Beauty and personal care: 5.39%
- Food and beverage: 4.80%
- Pet care: 4.71% (swinging between 2.24% and 7.68% month to month)
- Multi-brand retail: 3.01%
- Fashion, accessories and apparel: 2.77%
- Consumer goods: 2.47%
- Home and furniture: 1.22%
- Luxury and jewellery: 0.72%
Dynamic Yield does not publish electronics separately. The closest recent data for electronics is Triple Whale's paid-ad medians: over the same period electronics sits at the bottom with 1.49%, while food and beverage (2.60%), pets (2.39%) and beauty (2.38%) lead; apparel and accessories is at 1.80% and home and garden at 1.64%. The order runs the same way as Dynamic Yield's, the level is lower, because this figure measures only traffic from ads, much of which does not yet know the brand.
On the UK side, IRP Commerce's August 2026 data teaches the same lesson at small-store scale: fashion clothing and accessories 1.86%, health and wellbeing 3.31%, kitchen and home appliances 2.98%, food and drink 1.58%, baby and child 0.57%. For fashion, the band across three sources runs from 1.8% to 2.8%; a fashion store can place itself somewhere inside that band rather than against a single number.
Our reading: a sector table is for calibrating expectations, not for setting targets. A furniture store converting at 1.2% is not in trouble, it is in the middle of its sector; a beauty store running at the same rate is at around a quarter of its sector average. Reading sectors in Turkey needs two more corrections, return rates and basket size; we cover both below with figures.
How much lower is mobile conversion than desktop?
In most sources, clearly lower: in Contentsquare's 2026 report desktop visits convert at 3.4% and mobile visits at 2%, putting desktop 74% ahead; in retail the gap is 3.7% against 2%. Yet 69.9% of visits come from mobile, so most traffic is on the device that converts worse.
Littledata's 2023 Shopify data points the same way: desktop 1.9%, mobile 1.2%. Dynamic Yield's last 12 months say the opposite: mobile 2.88%, tablet 2.85%, desktop 2.37%. On top of that, Dynamic Yield's desktop rate fell from 3.35% in November 2025 to 1.76% in July 2026.
The most plausible explanation for the contradiction, in our view, is the sample's sector mix. By our average of the monthly device shares Dynamic Yield publishes, about 91% of traffic in beauty, the best-converting sector, comes from mobile; in home and furniture, one of the lowest, the share is around 63%. When the high-converting sector concentrates on mobile, the device average starts measuring the sector rather than the device. The lesson: compare your mobile rate not with your desktop rate but with your sector's mobile rate and your own history.
The device gap also shows where the funnel opens up. At Dynamic Yield, mobile visitors add products to the cart more often (6.35% against 5.21% on desktop) but also abandon the cart more often (79.84% against 69.48%). On mobile the problem is usually not interest but the checkout step: forms, card details and address entry on a small screen.
How much does traffic source affect conversion rate?
As much as sector does: in Contentsquare's 2026 data, paid search has the highest rate among paid channels at 2.8%, while organic social stays at 0.7%. A visitor who arrives by searching comes with a need; a visitor who taps a post in a feed usually has not decided yet.
The same report has two more findings. First, returning visitors convert at 2.9% and new visitors at 1.7%, and 52.8% of visits now come from returning visitors. Second, the conversion rate of traffic referred by AI assistants rose 55% in a year to 1.3%; its share is still small, but it is the only channel whose conversion is growing. That is why being visible in AI search is no longer only a brand matter.
For the level of paid advertising itself, Triple Whale's median is a good anchor: across more than 53,000 brands between August 2025 and July 2026, the median conversion rate of paid-ad traffic was 1.69%, down 4.63% on the previous period.
The practical consequence: when the channel mix changes, the total rate changes too. A drop in the month you raise the ad budget may not be a failure; new traffic that does not know the brand pulls the average down. Track each channel's own rate against that channel's benchmark rather than the total: comparing ad traffic with 1.69%, paid search with 2.8% and social with 0.7% tells you far more than comparing the total with 2.72%.
What is the average cart abandonment rate, and what do funnel steps show?
According to the average Baymard Institute calculates from 50 studies between 2006 and 2025, 70.22% of carts do not end in a purchase; in Dynamic Yield's last 12 months, 77.55% of products added to carts are not bought. Adding to cart is a signal of intent, not a promise to buy.
The studies Baymard compiles range from 55% to 84.27%; there is no single "normal" abandonment rate, there is a band. Part of abandonment cannot be prevented anyway: in Baymard's latest reasons study with US online shoppers, 42% had left a cart because they were "just browsing". Setting that group aside, the reasons rank as follows:
- Extra costs (delivery, tax, fees) too high: 40%
- Delivery too slow: 20%
- Did not trust the site with card details: 19%
- The site required an account: 18%
- Checkout too long or complicated: 17%
- Website errors or crashes: 17%
Most of the list concerns the checkout flow rather than the product, so it can be fixed. By Baymard's measurement, the average US checkout shows 23.48 form elements by default, while an ideal flow manages with 12-14; the institute calculates that better checkout design alone can raise a large e-commerce site's conversion rate by 35.26% on average. The five tactics for reducing cart abandonment are this list's practical counterpart.
Benchmarks at each funnel step
Splitting the funnel into steps instead of one conversion rate turns a benchmark into a diagnosis. Dynamic Yield's same 12-month series gives three separate indicators: add-to-cart after a product page view at 6.08%, abandonment of products added to cart at 77.55%, purchases per visitor at 2.72%. Because the three are measured in different units they cannot be derived from one another; each is the anchor for its own step. Add-to-cart varies by sector too: 9.56% in beauty, 6.24% in fashion, 3.67% in home and furniture, 1.76% in luxury and jewellery.
The reading rule is simple: if your add-to-cart rate is above your sector's but your conversion is below it, the problem is not on the product page but in the cart and checkout. In the opposite case, when visitors do not even add the product to the cart, the places to look are the product page, the price and the traffic itself.
Why does conversion rate fall as order value rises?
Because the larger the amount, the longer the decision, and the same buyer visits several times before purchasing: in Triple Whale's data, electronics with an average order value of 113 dollars converts at 1.49%, while food and beverage at 63 dollars converts at 2.60%. At Dynamic Yield, too, the bottom of the list holds two high-value sectors: home and furniture at 1.22% and luxury and jewellery at 0.72%.
The Ministry of Trade's 2025 data shows how large this gap is in Turkey. Average basket value by sector:
- White goods and small appliances: 10,513 TL
- Home, garden, furniture and decoration: 9,388 TL
- Medical, personal care and cosmetics: 3,062 TL
- Clothing, footwear and accessories: 2,910 TL
- Electronics: 1,894 TL
- Food and supermarket: 889 TL
The largest basket is more than eleven times the smallest. Comparing a white goods store with a 10,513 TL basket against the rate of a grocery site with an 889 TL basket makes no sense. If you have to compare different price bands, use revenue per visitor rather than the rate, that is, conversion rate multiplied by average basket: a store converting at 1% with a 10,000 TL basket earns 100 TL per visitor, a store converting at 4% with a 900 TL basket earns 36 TL.
Is there conversion rate data specific to Turkey?
We found no public Turkish conversion rate series with a published method; Turkey-specific data on this is limited. The Ministry of Trade's report, based on ETBİS records, measures volume, basket value, return rates and sales channel, not conversion per visit; the rates quoted in e-commerce platform providers' articles are either carried over from foreign sources or unsourced.
So for a store in Turkey, the conversion rate benchmark has to come from foreign series for now. The job of Turkish data is to correct that benchmark: which months should not be compared, in which sector the net rate differs sharply from the gross, which channel should stay out of the site rate. The Ministry's 2025 report, published on 12 May 2026, and TÜİK's 2026 survey provide enough material for that.
Volume and demand
E-commerce volume in Turkey rose 52.2% in 2025 to 4.57 trillion TL, with 5.94 billion transactions; e-commerce makes up 19.3% of all trade. According to TÜİK's 2026 survey, 60.0% of individuals aged 16-74 order goods or services online, up from 55.7% a year earlier. The share of mobile shopping is absent from official reports; the Ministry's report gives no device breakdown. In global series, mobile's share of traffic ranges from 69.9% (Contentsquare) to 75.75% (Dynamic Yield); since there is no verified figure for Turkey, we do not give one in this article.
Campaign calendar: November is a period of its own
E-commerce's share of all trade rises to 22.4% in November and falls to a 17-18% band in the summer months. In some sectors the swing is sharper: in white goods the share is 35.3% across the year but 61.4% in November; in electronics 38.6% against 49.4%; in clothing 25.4% against 32.4%. Comparing a month that includes November with one that does not means mistaking the campaign's effect for the site's success.
Sector: gross rate and net rate
According to the Ministry's data, cancellation and return rates in 2025 were 21.6% in clothing, footwear and accessories, 12.0% in electronics, 11.4% in sports and outdoor, 7.2% in white goods and 3.2% in medical, personal care and cosmetics. In clothing, roughly one order in five comes back as a cancellation or return. Foreign benchmarks count gross orders; a clothing store should calculate its rate both gross and net of returns, and check whether an improvement also raises returns.
Channel: marketplace sales stay out of the site rate
According to the Ministry's survey of 781 businesses, 48.8% sell both on their own site and on marketplaces, 39.5% only on marketplaces and 11.7% only on their own site. A marketplace order does not pass through your site's sessions; divide total orders by site sessions and your rate looks higher than it is. There is a Turkish correction on payments too: 62.5% of volume is paid by card, 29.2% by bank transfer and 3.5% cash on delivery. If the purchase event fires while a bank-transfer order is still awaiting confirmation, and cancelled transfers are never deducted, the rate is inflated again.
Should an SME, an exporter and a large retailer use the same benchmark?
No; if the stores producing the average are not at your scale, the number is no yardstick for you. An SME store should look at small-store series, a multi-market exporter at regional series, and a large retailer at enterprise platform series.
The SME store
The closest anchor is the distribution Littledata drew from 2,800 Shopify stores: an average of 1.4%, the top 20% above 3.2% and the top 10% above 4.7%. The data is from 2023, but because it gives a distribution rather than one average it is still useful: 1.4% means sitting in the middle, 3.2% means the top fifth. A more recent checkpoint is the 2.23% IRP Commerce measured in August 2026 for independent SME and mid-market stores in the UK and Ireland.
The real trap for a small store is statistical: in a store with 2,000 sessions a month, a difference of ten orders moves the rate by half a point. Work with a 90-day or 12-month rolling average instead of deciding on the monthly rate. In the survey of 108 sellers by iyzico, Dogma Alares and ETİD, 35% of sellers say they struggle with mobile compatibility and optimisation; given mobile's share of traffic, the mobile funnel should be the SME's first line of comparison.
The exporting, multi-market store
Conversion also varies by region: in Dynamic Yield's 12-month data, EMEA (Europe, the Middle East and Africa) is at 2.89%, the Americas at 2.66% and Asia-Pacific at 1.51%. Purchases by buyers abroad from sites in Turkey made up 3.7% of e-commerce volume in 2025; in the iyzico survey, 35% of sellers actively sell abroad and 44% name selling abroad and going global as a challenge. A multi-market store cannot be run on one total rate: in GA4 each country should be read as a separate funnel, and each market compared with its own regional average.
Baymard's list of reasons weighs heavier in exporting: extra costs (40%) grow with customs and international delivery, slow delivery (20%) with cross-border transit times. Our recommendation is to read a foreign market's rate not against your home market but against that market's regional series, together with how early the total cost appears in checkout.
The large retailer
For a large retailer, the Shopify average is the wrong anchor; Dynamic Yield's customer base of 200 million monthly unique users and Contentsquare's more than 6,500 sites are closer in scale. Being large does not mean the checkout is good either: Baymard's 35.26% improvement potential was calculated for large sites after testing the checkouts of sites such as Walmart, Amazon, Wayfair and ASOS. For a large retailer, the real job of a benchmark is internal comparison: setting category, device and channel breakdowns one by one against sector values and finding the one furthest away. For a large retailer that also sells on marketplaces, the site rate tells only part of the digital sales story.
How do you compare your own conversion rate with a benchmark?
First calculate your rate with the same definition as the source, then compare it over the same period and in the same breakdown; a comparison without these three matches misleads. In practice this means dividing sessions with a purchase by total sessions in GA4, taking a 12-month rolling average and reading device, channel and new-versus-returning breakdowns separately.
- Verify the purchase event. The purchase event in GA4 should fire once per order with the correct value; test orders, cancellations and bank-transfer orders awaiting confirmation should be filtered out. If the event is wrong, every later step is wrong.
- Choose and write down the denominator. A session-based rate (transactions ÷ sessions) is close to IRP and Contentsquare, a visitor-based rate to Dynamic Yield. State at the top of the report which one you use and compare only with a source that uses the same definition.
- Strip out the noise. Bot and internal traffic should be filtered; if the return from the payment provider's 3D Secure page attributes the purchase to that provider, its domain should go on GA4's unwanted referrals list, or the channel breakdown breaks.
- Match the period. In Dynamic Yield's global series the rate was 3.34% in November 2025, 3.30% in December and 2.33% in April 2026; do not compare a single month with a 12-month average. In Turkey, keep a separate comparison period for the November campaigns.
- Separate the breakdowns: device, channel, new and returning visitors, country. The total rate is a weighted average of these; when the mix changes, the rate moves without anything getting better or worse.
- Calculate the net rate as well. In sectors with high cancellations and returns, clothing above all, put the net-of-returns rate next to the gross rate.
- Compare with a band, not a single number. Write the values of two or three sources for your sector side by side and position yourself within that band.
Say a fashion store converts at 1.6% on a session basis. With the fashion band at 1.8% to 2.8%, the store sits slightly below it. But if the breakdown shows desktop at 2.6% and mobile at 1.2%, the problem is not the store as a whole but the mobile checkout step. The value of a benchmark lies in that second sentence: it tells you where to look.
Where do you start if you are below the benchmark?
Being below average is not a diagnosis by itself; first find at which funnel step and in which breakdown you fall behind, then fix only that step. Splitting the rate into funnel steps and setting each against the step values above will often tell you on its own where to start.
The first check should be the measurement itself. In the GYMWOLVES case the pixel and data flow were broken and ad decisions ran on incomplete data; in that state no rate could be compared with any benchmark. Once the data flow was repaired and the funnel rebuilt, sales rose 12× in three months. That figure is not a benchmark but one store's progress against its own starting point; still, it shows well why the order begins with measurement.
If the measurement is sound, use the seven critical pages method in the GAP analysis guide to see the gap page by page; the 21-tactic implementation catalogue lists what to do at each step. If you would rather start from what the rate is and how it is calculated, the article on what CRO is is a good introduction.
The first step needs no budget: scan your store's seven critical pages with Diagnoo; it also lists the gaps in your tracking set-up and shows how many points each gap costs. If you want to run the diagnosis together, our CRO consultancy starts with measurement and tests hypotheses in impact-effort order; the conversion rate optimisation service page sets out the steps and deliverables. If the problem is in the infrastructure, meaning the payment provider, marketplace integration or multi-market set-up, e-commerce consultancy handles that layer.
In one sentence: a benchmark is a compass, not a target. The test you can run today: split the last 12 months' session-based rate by device and channel, write each breakdown next to the value of the closest source in this article, and underline the row furthest away. That row is your next job.
Method and sources
This article is a compilation of public, anonymous and aggregated research data; no INDOLES client data was used. The figures were read from their sources on 25 September 2026; figures passed on second-hand that could not be verified at source were left out. We do not link out to the sources; you can find them by searching their names and dates.
- Dynamic Yield by Mastercard, eCommerce Benchmarks: monthly series and 12-month averages for August 2025 to July 2026; the company's customer base (200 million monthly unique users, 300 million sessions); conversion = completed purchases ÷ visitors.
- Contentsquare, 2026 Digital Experience Benchmark (25 February 2026): the final quarters of 2024 and 2025; more than 6,500 sites in 9 sectors, 99 billion sessions; conversion = share of visits that include a conversion event.
- IRP Commerce, eCommerce Market Data: August 2026; independent SME and mid-market stores in the UK and Ireland; conversion = transactions ÷ sessions.
- Triple Whale, Ecommerce Benchmarks 2026 (24 August 2026): August 2025 to July 2026; paid-ad medians across more than 53,000 brands.
- Littledata, Shopify conversion rate benchmark: 2023, 2,800 Shopify stores.
- Baymard Institute, cart abandonment rate list (last updated 22 September 2025): the average of 50 studies from 2006 to 2025; abandonment reasons from its study of US online shoppers.
- Republic of Türkiye Ministry of Trade, Türkiye'de E-Ticaretin Görünümü Raporu 2025, its annual e-commerce outlook report (12 May 2026): ETBİS data for 2025; the sales channel split comes from a survey of 781 businesses.
- TÜİK (Turkish Statistical Institute), Hanehalkı Bilişim Teknolojileri Kullanım Araştırması 2026, its household ICT usage survey (5 August 2026): individuals aged 16-74.
- iyzico, Dogma Alares and ETİD, Türkiye E-Ticaret Ekosistemi 2025, an e-commerce ecosystem report (17 June 2026): a survey of 108 sellers.
There is one contradiction between the sources and we leave it in the open: Contentsquare and Littledata measure desktop ahead, Dynamic Yield measures mobile ahead. The explanation in the device section, the sector mix, is our interpretation, not the sources'. The sector-level mobile traffic shares are likewise our own average of the monthly shares Dynamic Yield publishes.