Offer the add-on after the customer has chosen, never before. Until the main product is in the cart, every extra offer is one more question on the path to purchase.
The benchmark
Start with the distribution of your order values. A histogram in €10 bands shows where your orders cluster, and above all where there are almost none. It tells you whether a shipping threshold, a bundle or an add-on has a chance of working.
At a clothing brand the authors work with, nearly half of orders fell below the free shipping threshold, set at €120. The histogram showed two humps: small items between €50 and €90, and the flagship piece around €150.
Between €99 and €129, there was almost no one, even though the average order value landed roughly there, and so did the threshold.
Someone buying an €89 item had to add €31 to reach it, which meant a second item at the same price, something few people do to save on shipping costs. Someone buying the flagship piece cleared it without noticing.
The rule
Each step of the journey gets a single mechanism. On the product page, offer a higher-end version.
In the cart, suggest an add-on and remind shoppers of the shipping threshold. After payment comes one last offer, the only one that appears once the decision is already made. On Shopify, it runs through a post-purchase page app, available from the Basic plan.
The steps
- The distribution. Plot a histogram of your order values over three months, leaving out sale weeks (soldes) that skew the curve.
- The gaps and the add-ons. Spot the busiest band and the empty ones, then note the price of what your customers already add on their own.
- The threshold. Set it just above the main hump, within reach of a real add-on. If reaching it takes a second item at the same price, it's too high (the guide on the free shipping threshold walks through the calculation).
- The measurement. Track revenue per visit and the number of orders in addition to average order value, comparing against a control version or the same period last year. One customer put it well in a meeting: "you need something to compare yourself to, because otherwise, yes, of course, giving things away will always look better than not giving anything."
- The line for leadership:
"Average order value rose by this much, the number of orders moved by this much, revenue per visit by this much: that last figure tells us whether we won."
The trap
Average order value can rise because you lost the small orders. Raise the free shipping threshold, and small carts go elsewhere: the average climbs, the number of orders drops, and the dashboard reports a success.
It can also fall because of the very driver meant to lift it, and the authors learned that the hard way.
At a brand with a high average order value, they ran a free shipping from €300 test, with no doubt about the outcome. Add-to-cart rose 2.6%, conversion fell 24% and average order value dropped 36%. Nothing has been proven since, but one hypothesis is anchoring: the displayed amount became a reference price, and carts that used to build up beyond it stopped at the displayed amount.
A shipping threshold is still a price you display, and average order value should never be read without the number of orders next to it.
