SMS and Email Cadence Template

Cadence Map

One ordinary week: two email campaigns to everyone, one to the engaged segment, browse abandon, a three-message cart series, two post-purchase and a replenishment reminder. On SMS: one campaign to everyone consented, one to VIPs, and a cart abandon text.

Compare D1 and E1. Same activity, same programmes, one of them consented to texts. 13 messages against 10, which is 1.30x. 23.11 units against 10.00, which is 2.31x. The message count understates the load on the dual-channel group by nearly half, and the message count is what every platform reports.

The SMS list is not a smaller email list. 40.1% of SMS subscribers are VIPs against 9.0% of email subscribers, a 4.45x concentration, because people opt into texts from brands they already like. So the load falls hardest on the most valuable cohorts, and 88.5% of the SMS list is on email too.

Median load is 2 messages and 2.00 units; p99 is 10 and 20.11. The two orderings disagree, which is the whole reason to keep both columns.

Frequency Cap

One budget in attrition units, swept rather than chosen. Each person's units are spent on email first. The last column divides the cost of a saved contact by what that contact returns a year.

Cost per saved contact is nearly flat, $39.76 at the tightest budget and $36.08 at sixteen units. That is because the cut always lands on the marginal message, whose revenue is roughly the same wherever you make it. So the level cannot be argued from cost, which is what almost every cadence recommendation tries to do.

What moves is what a save is worth. An email subscriber returns $12.90 a year and an SMS subscriber $22.68, so a tight cap that starts cutting email is saving the cheaper contacts. Divide and you get a break-even tenure: ten units pays for itself at 1.59 years, two units needs 2.06. Put your retention curve beside that column.

A text earns 3.51x an email per send and 0.77x an email per unit of attrition, so under a shared budget the text is what gets cut at the margin. At ten units, email-first gives up $416 a week to save twenty more contacts, $20.32 each, against a contact worth $25.81 to $45.35 over two years. At fourteen units email-first wins on revenue as well.

Only 4.2% of the base is over a ten-unit budget today, which is 2,652 people in four cohorts. SMS is additive rather than substitutive for the other 91.4%, and nowhere else.

Consent Register

One row per consent event per channel, written at capture rather than reconstructed later. Seven rows shown, covering the four states that actually occur.

The p-38204 row is the one that costs money. The reply read “no more texts please”, which no keyword filter catches. A revocation made in any reasonable manner has to be honoured, a sender may not designate an exclusive means of revoking, and the window is a reasonable time not to exceed ten business days. If nothing reads free-text replies, that row is sitting unhonoured in an inbox.

The p-51993 row carries a consent state imported from a prior vendor with no disclosure and no signature record. The register marks it unsupportable and suppresses it rather than carrying a flag it cannot evidence. An imported list looks exactly like a consented list in every platform, and the difference only appears when somebody asks for the record.

The last row is a correct state that renders as an error in any dashboard keeping one subscription flag per person: SMS live, email revoked. Honoured by is the audit column, and a blank one on a revoked row is a finding on its own.