Every growth playbook says the same thing: find your hero products and put everything behind them. We opened 31.5 thousand orders to see what heroes actually do. They carry the revenue, open the door to new customers and fill bigger baskets. They do not, on their own, bring customers back.
A hero product is the product a store is known for: the one that brings in an outsized share of revenue, leads the ads and the home page, and is often the first thing a new customer buys. In skincare it is the signature serum; in fashion, the cut everyone asks for.
In marketing, the hero is the product a brand leads with. Most store owners mean something more practical: the few products that pay the bills. That is how we defined it: in each store, the top ten products by revenue over the period. A bestseller list ranks by units. A hero list ranks by money.
The usual advice is to find your heroes and push them hard. Some of it holds up in the data, and some does not. The data is aggregated from online stores we grow: 31.5 thousand orders, January to September 2026. Every figure below is a share, a ratio or a median, so the lessons travel to any market.
Revenue concentrates fast. In the median store, the single best-selling product brought 6% of item revenue. The top three brought 16%, the top five 25%, the top ten 39% and the top twenty 54%. The median store sold about 230 different products, so a small slice of the catalogue carried well over a third of the money.
At the other end, about 68% of the products that sold at least once sold fewer than ten units in nine months. Most of a catalogue is long tail.
The median hides a wide range, though. Across the stores we measured, the top ten brought anywhere from 11% to 89% of item revenue. A store with hundreds of similar products and a store built around one signature item are different businesses. Measure your own concentration before you borrow anyone's benchmark.
In the median store, 35% of all orders included at least one top-ten product. Among first orders, the first purchase we saw from each customer, the share was 36%.
Read it both ways. Ten products out of about 230 appear in more than a third of first orders, which is a lot of work for a short list. And yet 64% of first orders had no hero in them at all. Heroes are the busiest door into the store, not the only one, so the rest of the catalogue still deserves decent pages.
In the median store, an order with a hero product was worth 1.38 times an order without one, about 38% more. Store by store the ratio ran from 0.88 to 3.01, so in some stores hero orders were actually smaller.
The tempting story is that the hero pulls the catalogue into the basket. The data supports that only in part. In 25% of hero orders, the customer also bought a product outside the top ten. The other 75% stopped at hero products. The bigger basket probably mixes several things: heroes that cost more than the average item, customers buying more than one, and sometimes a second product. This data cannot separate them.
That is the opportunity. A customer with a hero in the cart is already spending more, and most leave without anything else from the store. What a well-chosen second item adds, and how to offer it, is in how to increase average order value.
This is where the common advice breaks. A great hero is supposed to win a customer for life. In our data, customers whose first order included a top-ten product came back a little less often, not more: 7% of them ordered again within the period, against 9% of customers whose first order had no hero.
It is not the same everywhere. Store by store, customers who started with a hero came back at anywhere from 0.41 to 1.75 times the rate of the others. In the median store, though, the hero gave no loyalty advantage.
Why might that be? The data does not say, so these are hypotheses, not findings:
A caveat on the window, in plain words. Our data starts in January 2026, so the first order we see is not always the customer's first order ever. Customers who first bought in late summer had only a few weeks to come back. Both effects pull repeat rates down on both sides, so read them as a floor, and trust the comparison more than the level.
The hero brings the first order. The second order is built by what happens after it.
So keep pushing heroes, but stop expecting them to do the retention work. The second order comes from what follows the first: post-purchase emails, a bundle that introduces a second product, a clear next product to try.
Hero products are less permanent than the word suggests. In the median store, only 50% of the top ten from January to April was still in the top ten from May to September.
Seasons move demand, promotions lift products for a few weeks, launches take the top spots and stock-outs pull others down. A hero list made once soon describes a store that no longer exists. Re-check it every quarter, and again before every big sale season.
A common assumption is that heroes sell at full price and discounts are for the rest. In the median store, 53% of hero revenue came from orders that carried a discount, against 61% for other products. Compared within each store, the typical gap between the two was 0.2 percentage points. In practice, heroes sold on discounted orders about as often as everything else.
One caveat: a discount is counted at the order level, so it may have applied to another item in the order. The point stands: heroes live inside the promotion calendar, not outside it. If you shield a hero from every promotion, check that you are protecting a margin and not a habit.
You need one export and an afternoon:
If you would like a second pair of eyes on your hero list, we are glad to talk.
A hero product is the product a store or brand is best known for and leads with. It brings in an outsized share of revenue, carries the ads and the home page, and is often a new customer's first purchase. In our data we defined the hero set as a store's top ten products by revenue: in the median store they brought 39% of item revenue.
Export your line items for the last six to twelve months, group variants into one product, rank by revenue and add a running share of the total. The products before the curve flattens are your heroes. Then check how often each one appears in first orders and whether those customers come back, and repeat the exercise every quarter.
There is no fixed number. In the median store we measured, the top three products brought 16% of item revenue and the top ten 39%. A practical split is a few heroes to lead the ads and the home page, and a list of about ten to watch, keep in stock and re-check every quarter.
Not by themselves, in our data. Customers whose first order included a top-ten product ordered again at 7%, against 9% for customers whose first order did not. The data is observational and covers nine months, so treat the rates as a floor. The hero wins the first order; emails, bundles and the next product win the second.
A bestseller is ranked by units sold. A hero product is defined by its role: the revenue it brings and the customers it attracts. The two often overlap, but a cheap accessory can top the units list without being a hero, and a premium signature item can be a hero with modest unit sales. Rank by revenue when you decide where attention and budget go.
Most advice on raising average order value is a list of tactics. We opened 42,570 orders from online stores to see what the basket actually looks like, and which levers the data really supports.
Daniel Hirsh29 September 202610 min readEcommerce is easy to define and harder to run. This guide covers what it means, how an order travels from the first click to the doorstep, the main types of ecommerce, what a business needs behind the store, and what 42.6 thousand real orders show about how online stores make and grow their revenue.
Daniel Hirsh8 October 20269 min readMost guides to building an online store start with picking a theme. We start with the data: 42.6 thousand orders in online stores we grow show where sales come from, when they arrive and who comes back. The build order follows from that: which pages first, how many products to launch with, what to set up on day one, and what it costs in money and time.
Daniel Hirsh8 October 202610 min readEvery Preneuriat client gets a live dashboard: store revenue, all media spend, real ROAS, updated all day. The first call goes both ways.
Check the fit
