Abandoned cart: recovering the sale without nagging

Most people who put something in a basket and leave were never going to buy that day — they were comparing, or interrupted, or checking the delivery cost. A small share of them will buy if reminded, and recovering that share is one of the few reliably profitable automations in online selling. It is also one of the easiest to overdo, because the mechanism rewards sending more and the customer relationship does not.

What an abandoned cart sequence should contain

• A first message that is a reminder rather than a pitch — the items, a clear way back, and nothing else.
• The information that probably caused the hesitation: total delivery cost, delivery time, return terms.
• A second message a day or so later, if at all, acknowledging that the first one may have been ignored for a reason.
• A clear and immediate way to stop receiving them.
• Not a third, fourth and fifth, which is where recovery turns into harassment and unsubscribes outrun orders.

Timing, which matters more than wording

The first message belongs within a few hours, while the decision is still live and the person remembers what they were looking at. Beyond about a day the context is gone and the message reads as surveillance rather than service. The exception is anything with a long consideration cycle, where a reminder after several days can genuinely help — but that is a different message, closer to "still thinking about this?" than to "you left something behind".

You need a lawful basis to email somebody who did not complete a purchase. An address captured mid-checkout was given to buy something, not to be marketed to, and in several jurisdictions a recovery message is fine while a promotional one is not. Where the line sits depends on where your customers are, and the safe version keeps the message about the specific basket rather than about your range.

When a discount costs more than it recovers

Almost always sooner than expected. Offer ten per cent in the first recovery message and you quickly teach regular customers to abandon deliberately — the discount becomes a price cut applied to people who would have paid in full, which is the opposite of recovery. If a discount is used at all, it belongs late in the sequence, at a modest level, and ideally only for first-time buyers. Free delivery is usually the better lever because it answers the actual hesitation rather than reducing the price of everything.

Measuring it honestly

1. Count recovered orders against messages sent, not against carts abandoned, since most of those were never sales.
2. Subtract the orders that would have happened anyway — people who return by themselves are counted as recoveries by most systems.
3. Watch unsubscribes per sequence, which is the cost side and the one platforms report least prominently.
4. Compare revenue recovered with margin given away where a discount is used; recovery at a discount can be a loss presented as a win.
5. Check the sequence quarterly for stale content: nothing dates faster than an automated message nobody has read in a year.

Ettex Email handles the sequence itself with the basket contents and a plain way back, and the honest judgement comes from the conversion tracking beside it rather than from the platform's own recovery figure. The hesitation it is answering usually started earlier, on the checkout page, which is where a fix removes the problem rather than mitigating it.