Bundles that people actually buy

Bundling is old and mostly done badly. The usual version — three arbitrary products, ten per cent off, presented as a favour — does not work, and it does not work for a reason worth understanding.

What a good bundle is

A bundle should remove a decision rather than add one. The shopper’s question is not “do I want a discount”, it is “what do I need to make this work”. A bundle that answers the second question sells. One that only answers the first is a coupon with extra clicks.

The test: could a knowledgeable shop assistant have suggested this set, with a straight face, to somebody buying the main item? If not, it is not a bundle.

The three that reliably work

• The complete-the-job bundle. The thing plus what it needs — the frame with the print, the filter with the jug, the blades with the trimmer. Highest attach rate of any format because it prevents a second trip.
• The routine. Products used in sequence, shown in sequence. Video is unusually good at this: the order is visible, which is most of the explanation.
• The set with a real saving. Genuinely cheaper than buying separately, by enough to notice. Five per cent is noise. Fifteen is a reason.

The discount has to survive the shopper

This is the part that goes wrong technically rather than strategically. A bundle discount drawn by the widget — a number on a page, applied by JavaScript — does not survive a shopper who adds the bundle, closes the tab, and checks out from their phone two days later.

What they see at checkout is the undiscounted price and what you get is an email about it. The discount has to be a real Shopify discount, applied at checkout by Shopify, so that it holds across sessions and devices.

One tap

If the shopper has to add three items individually, you have not built a bundle, you have built a list. The whole value is that a set of decisions becomes one decision and a set of taps becomes one tap.

Let people buy one thing

The most common way to over-engineer this: making the bundle the only route to the products in it. Somebody who wants the jug and not the filters should be able to buy the jug without hunting for it.

The bundle is an offer, not a gate. Treating it as a gate costs you the single-item sales you were already making, and those are most of them.

When part of it is out of stock

Decide what should happen before it happens, because it will. A bundle that silently fails at checkout is worse than one that says plainly which item is unavailable and keeps its price visible. The shopper can then decide, which is all they wanted.

Measuring it

Average order value on sessions that saw a bundle, against sessions that did not. Not bundle conversion rate in isolation — a bundle that converts at two per cent and lifts the average basket by forty per cent is working. Judged on conversion alone it looks like a failure.