CRO Diary · a conversion specialist's working notebook entries reviewed weekly
CRO Diary

More sales from the traffic you already have

Experiments

How to Increase Average Order Value on Shopify

The average Shopify order is 85 dollars. Here is how I design AOV experiments that actually move that number, the mechanics behind each tactic, and the trap that makes most AOV wins fake.

From the diary · CRO Diary  ·  July 16, 2026  ·  6 min read  ·  AI-drafted from sourced data

A desk with printed store receipts, a calculator and an open notebook with handwritten figures
Photo via Openverse (CC0)

Average order value is the metric owners come to when they have made peace with their traffic bill. Getting more visitors costs money every single time, but getting the visitors you already have to buy a little more costs a design decision, once. That arithmetic is why AOV experiments fill so many pages of my diary, and why this entry collects the tactics I keep returning to, with the mechanics of why each one works and the measurement trap that makes half of all AOV “wins” fake.

Know your number and its neighborhood

Littledata’s benchmark of 2,800 ecommerce sites gives the reference points. The average order value across ecommerce was 101 dollars, and for Shopify stores specifically it was 85 dollars. More than 192 dollars puts a Shopify store in the top 20%, more than 311 dollars in the top 10%, and less than 54 dollars across ecommerce lands in the bottom 20%.

Before running anything, pull your own AOV two ways: the mean, which is what dashboards show, and the median, which is what most of your customers actually do. A store with a 90 dollar mean and a 40 dollar median is really a 40 dollar store with a few whale orders, and the tactics below play very differently against those two shapes.

THE AOV RULER · WHERE ORDERS LAND LITTLEDATA BENCHMARK, 2,800 SITES $54 BOTTOM 20% $85 SHOPIFY AVG $192 TOP 20% $311 TOP 10% ECOMMERCE OVERALL AVG $101 · SOURCE: LITTLEDATA.IO
The ruler I draw before any AOV experiment. Your median order tells you which tactics are even in reach.

Experiment 1: the free shipping threshold

This is usually the first experiment I run on a store without one, because its engine is the strongest documented force in cart psychology. In Baymard’s survey of abandonment reasons, 39% of US shoppers had abandoned a cart because extra costs, led by shipping, were too high. People do not merely dislike shipping fees, they resent them, and a threshold converts that resentment into motivation: add one more item and the fee you hate disappears.

The mechanics that matter. Set the threshold moderately above your current typical order, so it is reachable with one added item, not two. Announce it early, on the product page and in the cart, not as a checkout surprise. And show the remaining distance in the cart, because “you are 12 dollars from free shipping” is a call to action while “free shipping over 75 dollars” is a poster. The failure mode is a threshold set so high that customers ignore it, in which case you have simply told everyone that shipping costs money.

Experiment 2: bundles that already exist in the data

The bundles that work in my experience are never invented, they are discovered. Somewhere in your order history there are pairs of products that keep being bought together, or bought two weeks apart by the same customer. A bundle formalizes that pairing, prices it slightly kinder than the sum, and puts it where the first product’s momentum can carry it.

The mechanics: the discount should be visible but modest, because the real value of the bundle to the customer is decision relief, not the 10%. And the bundle must answer an obvious “because”: camera because memory card, planner because the pen that fits its loops. When I see a bundle that needs a paragraph to justify the combination, I already know how the experiment ends.

Experiment 3: the cross sell at the moment of yes

Placement decides whether a cross sell reads as service or as pestering. The moment right after adding to cart is the best real estate in the store: the customer just said yes, the cart is open, and a single relevant suggestion, accessory, refill, protection, rides that yes. The same suggestion shown as a popup before the customer has chosen anything is an interruption, and shown at checkout it is a distraction from the one flow you never want to complicate.

My rule for the experiment: one suggestion, visibly cheaper than the main product, explainable in one line. Stacking four recommendation carousels between cart and checkout raises the theoretical basket and, in every log I have kept, taxes the completion of the real one.

Experiment 4: quantity logic for consumables

If your product runs out, socks, coffee, filters, supplements, the cheapest AOV win is making the multi unit choice legible: a two pack and a three pack with per unit prices spelled out, next to the single. The mechanic is not discounting, it is arithmetic done for the customer. Most buyers of consumables are willing to buy ahead, they just never do the division. Stores selling one-at-a-time consumables are leaving the easiest experiment on this list unrun.

The trap: AOV that eats conversion

Now the part that earns this entry its place in the experiments section. Every tactic above can produce a beautiful AOV chart while quietly making the store poorer, because AOV has an evil twin: order count. Push thresholds too high, stack upsells too thick, and the average order grows while the number of orders shrinks. The tide goes out, the remaining boats are bigger, the harbor is emptier.

An AOV experiment is never “did the average order grow”. It is “did revenue per visitor grow while conversion held”. Any result that cannot answer the second question is an anecdote with a chart.

So every AOV experiment in my diary carries three lines, not one: AOV, conversion rate, revenue per visitor. The third is the referee. And one supporting fact from the speed research belongs here, because it surprises people: in the Milliseconds Make Millions study published on web.dev, a 0.1 second improvement in mobile speed metrics was associated with retail customers spending 9.2% more per order. Basket building is browsing, browsing is page loads, and slow pages tax exactly the wandering that fills carts.

Where the potential sits

The gap between an 85 dollar average and the 192 dollar top 20% line is not closed by one weapon, and it is mostly closed by stores whose products allow it. But in the audits I run, most stores are operating none of the four experiments above, not even the shipping threshold. Running just the first two, honestly measured, is usually worth more than any redesign conversation, because every dollar of AOV you add is earned again by every order that follows, at zero additional traffic cost. That compounding is the whole reason this metric deserves its own page in the diary.

Questions I get about this

What is the average order value for a Shopify store?

Littledata's benchmark of 2,800 ecommerce sites puts the average Shopify order at 85 dollars, against 101 dollars for ecommerce overall. More than 192 dollars would put you in the top 20% of Shopify stores, and more than 311 dollars in the top 10%.

What is the fastest way to increase AOV on Shopify?

A free shipping threshold set moderately above your current average order is usually the first experiment I run, because the mechanics are backed by Baymard's finding that extra costs are the number one reason shoppers abandon. It uses a fee people already hate as the engine of a bigger cart.

Do product bundles increase average order value?

They can, but only when the bundle solves a real pairing the customer already makes, like a device plus the accessory everyone buys two weeks later. Bundles invented in a spreadsheet to hit a price point tend to fail because the customer sees no reason for the combination.

Can increasing AOV hurt my conversion rate?

Yes, and this is the most common failure I see. Aggressive upsells, popups, and padded thresholds add friction and cost, so the average order grows while the number of orders drops. Always read AOV and conversion rate together, and judge experiments on revenue per visitor.

Sources & data

  1. Littledata, average order value benchmark from 2,800 ecommerce sites (2023 benchmark, accessed July 2026)
  2. Baymard Institute, 50 Cart Abandonment Rate Statistics, reasons for abandonment (updated September 2025, accessed July 2026)
  3. web.dev, Milliseconds make millions: impact of 0.1s speed improvements on retail spend (accessed July 2026)
Cite this entry: CRO Diary (2026). “How to Increase Average Order Value on Shopify.” https://crodiary.com/experiments/how-to-increase-average-order-value-shopify/