Journeys: the campaigns that run whether or not anyone is at their desk

Broadcast campaigns are the work you do. Journeys are the work that gets done for you. A location running seven well-built journeys generates a substantial share of its marketing-attributable revenue without anyone touching a keyboard that week, and because journeys fire on customer behaviour rather than on the calendar, they arrive at the moment the message is relevant.

Build them in this order

Do not build all seven at once. Each one takes a month to prove out, and the order below is by return on effort.

1. Second order. Triggered 3–7 days after a first ever order. Attacks the steepest drop-off in the file. Almost always the highest-return automation a pizza shop can build.
2. Welcome. Triggered on signup without a purchase. Two or three messages establishing who you are, what to order first, and how to get it fastest.
3. Drift. Fires at 1.5× the customer’s median interval. No discount: a reason, not a bribe. Cheap and surprisingly effective.
4. Win-back ladder. Three escalating touches from the lapse threshold. Where real budget goes.
5. Birthday and anniversary. High open rates, high goodwill, trivially easy to run, and the anniversary of a first order is under-used compared with birthdays.
6. Habit builder. For customers at 2–4 orders, aimed at establishing a weekly or fortnightly rhythm rather than at any single order.
7. Post-catering / large order. A large order is a lead, not a transaction. The follow-up should treat it that way.

The second-order journey in detail

Because it matters more than the rest combined, it is worth spelling out. A first-time customer has bought once and has no habit. The majority will never order again, and the ones who do become dramatically more valuable.

Entry, exit and suppression rules

Most journey failures are not creative failures. They are logic failures: a customer stuck in two journeys, or receiving a win-back the day after ordering.

• One journey at a time. Entering a journey should suppress entry to lower-priority journeys until exit.
• Exit on conversion, immediately. If the customer orders, the journey ends on that event, not at the next scheduled step.
• Exit on channel loss. An SMS opt-out ends the SMS arm of every journey instantly, and that must be enforced at send time rather than at build time.
• Re-entry cooldown. A customer who completed a win-back ladder should not re-enter it for at least twice the ladder length, or you have built a machine that harasses your least engaged customers.
• Global cap applies. Journeys are high priority but not exempt from the total contact cap.

Journeys need holdouts too

A journey that has run for a year without a control group has a year of unfalsifiable reporting attached to it. Hold out 5–10% of eligible entrants permanently, and read the difference in order rate and revenue per customer rather than the journey’s own click-through.

This regularly reveals that one of the seven journeys is doing nothing. Finding that out is worth more than the journey was.

An automation you cannot switch off to test is an automation you cannot defend.

Questions