Orders averaged, not customers
Average order value, written AOV, is the typical amount a shopper spends in a single transaction on your store. Work it out from revenue and order count over a fixed period, and read it as a ceiling rather than a score: it caps what you can afford to pay for a sale, because the gross profit inside one order is all you have to cover acquisition. Raise it and a previously unprofitable click becomes affordable.
It counts orders, not people, which trips up a lot of dashboards. One customer buying three times in a month appears as three orders, so a brand with strong repeat purchase can show a modest average and still earn well per shopper. If the per-person figure is what you actually want, you are looking for revenue per customer or lifetime value. Decide too whether your version sits before or after discounts, shipping and returns, because only the net one is money you can spend.
The mean also hides shape. A handful of large orders can lift it while most baskets sit within a few dollars of the entry price, and a single wholesale transaction can move a whole month. Glance at the distribution, or at the median, before you build a plan on the average. If half your orders are the cheapest item in the range, that is the number your advertising is really up against.
How bundling changes the maths
Basket size answers to merchandising first. Bundles, multipacks, a subscription option, free shipping above a threshold and one relevant upsell at checkout all move it, and none of them require a new ad. Say the entry product sells at thirty dollars and a three-pack at seventy-five. Move a quarter of orders into the three-pack and the average rises by several dollars on the same traffic. Treat that as the mechanic rather than a forecast.
Now follow it through to what you can pay for a customer. At a forty percent margin, a thirty dollar order leaves twelve dollars to cover acquisition; a seventy-five dollar order leaves thirty. The same click price that was losing money at the entry basket is comfortable at the bundle, which is why a bundling push often does more for cost per acquisition targets than another round of audience testing.
The lever creative can actually pull
Creative moves basket size only where the offer already exists. A video showing the product used twice a day makes a multipack feel like the sensible choice. A morning routine sells three items where a single-product demo sells one. A restock moment sells the subscription. If nobody has built the bundle or set the shipping threshold, no opening line is going to invent one for you.
So write the basket into the brief rather than hoping for it. Ask for the bundle in frame instead of the hero product alone, and for the creator to explain in their own words why they keep more than one around. Shelf shots, a full routine, a spare unit in the cupboard and an offhand line about running out do more for order value than another slow close-up of the packaging.
Judging creators on basket size
Then read the figure per creative, not just per account. If one creator's videos bring a higher basket at a similar click price, they are worth rebooking even when their view counts look ordinary. Basket size is the quiet half of return on ad spend: revenue over spend improves either by paying less for attention or by selling more per order, and the second route is usually the one nobody is working on.
Two cautions before you act on it. Attribution at basket level is noisy on small volumes, so give each video enough orders to mean something. And a rising average is not automatically good news: if it climbed because your cheapest product went out of stock, you have a supply problem dressed up as a win. Check what changed in the mix before you credit the creative.
How it's used
Average order value on the bundle page is running about eighteen dollars above the single product page, which is why the next two creator videos should be built around the three-pack.
For this shoot, please show the full set on the counter and mention that you go through a bottle roughly every three weeks. We are selling the multipack here, not the sample.