DEV Community

Swipe Again for Swipe Again

Posted on

A Text Gets Read. That Is Exactly Why It Must Be Rare.

Email dunning has a quiet problem that no subject line fully solves: a large share of it is never opened. The renewal failed on Tuesday, the first email lands in a promotions tab, the second is skimmed past on Thursday, and by the time the customer actually sees one, the account is a day from cancellation and the fix feels like a rescue instead of a routine correction. The message was fine. The channel was where it went to wait.

A text message behaves differently. It is read, usually within the hour, by someone holding the device the message arrived on — which is also the device they can fix the problem from. That combination, speed plus proximity to the fix, is what makes SMS valuable in payment recovery. It is also what makes it dangerous to waste. A channel people actually read is a channel people will mute the moment it starts behaving like marketing.

So the discipline is scarcity. One text at the moment that matters — the first failure, while the fix is still a thirty-second job — outperforms a drip of reminders across two weeks. The text should carry three things and nothing else: what happened in plain words, that the service continues while it is sorted, and a single link that opens the card update directly, with no login wall and no detour through a homepage. Every additional sentence spends attention the next text will need.

Timing is the second discipline. A text sent at 2 a.m. because the retry batch ran at 2 a.m. announces that the business does not think about the person receiving it. Recipient-local hours — morning to evening, in the customer's own timezone — are not a courtesy detail; they are the difference between a helpful nudge and an intrusion that gets the number blocked, which ends the channel for that customer permanently. Quiet hours are a compliance matter in many places besides, and treating them as a design constraint from the start is cheaper than retrofitting them after a complaint.

Measure it separately, too. A text's recovery rate should never be blended into the email sequence's numbers, because the two channels reach different customers at different moments — the text often gets credit for saves the email had already softened up, and email gets blamed for failures it never had a chance to reach. Report the channel honestly: how many first failures received a text inside the first hours, how many of those updated a card from that link, and how many simply replied and were helped by a person. That last group is easy to overlook and worth naming — a text is two-way in a way email rarely is, and some customers will answer it with a question rather than a click. Staffing that reply path, even lightly, converts confusion that would have become a cancellation.

Consent and exit belong in the same conversation. The customer agreed to be contacted about their account; a failed-payment text is squarely inside that agreement, and a marketing offer smuggled into it is not. And every text must honour an opt-out instantly and permanently for that channel, shifting the sequence back to email without argument. A customer who texts STOP has not churned — they have stated a preference, and respecting it keeps every other door open.

Used this way, SMS is not a louder email. It is a different instrument: the one message in the sequence that is meant to be read now, by the person who can act now. The businesses that get value from text reminders that arrive while a fix is still possible are the ones that protect the channel's credibility as carefully as they use its reach — few messages, right hours, one link, no noise.

That restraint is built into how Swipe Again runs recovery: text used sparingly inside recipient-local hours, email carrying the longer sequence, card updates one click deep, and recovery counted on the original dollars at a flat $179 a month. The text gets read because it is rare. Keep it rare.

Top comments (0)