AOV, conversion or retention: which Shopify lever to pull first

Every Shopify growth conversation lands on the same three levers. Convert more of the visitors you already have. Get each order to be bigger. Get customers to come back. Each one is a real business, with its own apps, agencies and advice. The problem is choosing, because a small team can only run one at a time properly. This article gives you a simple way to price each lever in your own store's dollars, and the order we recommend. The benchmarks come from Littledata's October 2026 panel of 433 Shopify stores: median conversion 1.4%, median order value $111.^[1]

The model

Annual revenue is sessions, times the share that buy, times the average order, times how many orders a customer places. Take a store at $1M a year with the Littledata medians: 1.4% conversion and $111 per order.^[1] That is about 9,000 orders from roughly 645,000 sessions. Assume 1.25 orders per customer per year, which is typical when a store has not worked on retention. Now move each lever by a realistic amount and hold the others still.

Three things stand out. The conversion move is the biggest, and it is only three tenths of a point: the median store moved about a quarter of the way toward the top-20% mark of 2.7%.^[1] The AOV move looks small but $12 is a 10% lift, which is what a well-placed bundle or free-shipping threshold typically produces. And the retention move needs one extra order from every fifth customer, which is a real programme, not a setting.

Why conversion goes first

Not because it is biggest, although in the example it is. Because it is cheapest and it compounds. Most conversion fixes are removals: a shipping cost that appears late, a forced account, a slow product template, a missing size chart. They cost days, not months, and once fixed they stay fixed. And every visitor you convert is a customer that AOV and retention work can then act on. Raising order value on a leaking funnel increases the value of the carts you lose. We cover where to look in the funnel benchmarks article.

The one exception: if conversion is already above the top-20% mark for your category, you are past the cheap fixes, and the next tenth of a point costs more than the first $12 of order value. Move on.

Then order value

AOV is a pricing and merchandising problem, and it responds to a small number of moves: a free-shipping threshold set just above the current median order, a bundle of the two products most often bought together, and a size or quantity option that is better value per unit. Each is one test. The Littledata panel shows the spread: median $111, top 20% at $290.^[1] That gap is mostly category, not skill, so benchmark against your own history rather than the panel.

Watch the median order, not the mean. A handful of wholesale or bulk orders will drag the mean up and hide a flat result for ordinary customers.

Then retention

Retention is the most valuable lever over years and the slowest to show. It needs a product people re-buy, an email and SMS programme, and often a subscription option; which products qualify is its own question. It also needs clean customer data, which is why it comes after the tracking and funnel work. The payoff is that a repeat customer costs nothing to acquire, and Shopify's 2026 data gives one more reason to care: shoppers arriving from AI assistants carry 14% higher average order values and include 1.3 times as many first-time customers, so the front door is bringing in more new people than it used to, and keeping them is where the margin is.^[2][3]

The one number per lever

• Conversion: checkout completion rate, weekly, by device. It moves first when a fix works.
• AOV: median order value, weekly, excluding wholesale.
• Retention: share of customers who place a second order within 90 days, by first-order month.

Put the three in one sheet with the model above, and every proposed change can be priced before it is built. That is the whole discipline: fewer tests, each one with a dollar figure attached.