How a delay works
When somebody reaches a delay in a sequence, the platform parks them there and holds a timestamp. When the wait is up, they move to the next step. Nothing else happens in the meantime, but the person is still inside the flow and still counted in its reporting.
Most platforms offer two kinds. A relative wait pauses for a set stretch of time measured from the moment the person arrived at the step. An absolute wait holds them until a particular clock time, day of the week or date arrives, which is how a sequence stops sending at three in the morning. Some tools add a third form that waits until a condition becomes true, such as an order being marked as delivered.
The time zone that a relative wait resolves against is a setting, not a given. It may be the account’s zone, the subscriber’s own zone if the platform has recorded it, or the server’s — and getting this wrong is the reason well-written sequences arrive at odd hours.
Why delays matter
The gaps are what turn a set of messages into a sequence. The same emails sent in quick succession feel like pestering; spread sensibly they feel like a business that is paying attention. Nothing else in a flow changes the tone as much for as little effort.
Delays also give reality time to catch up. A review request needs the parcel to have arrived. A follow-up to a quote needs the client time to read it. A branch that checks whether somebody has bought needs a gap long enough for buying to have been possible, or it will send everyone down the same path.
Where delays go wrong
The most frequent fault is a relative wait that lands outside waking hours. A signup late at night, plus a wait measured in hours, puts the second message on the customer’s phone while they are asleep, and a business email that buzzes at dawn is remembered for the wrong reason.
The other is drift. Each step looks reasonable on its own, then somebody adds a branch with its own wait, and the sequence quietly stretches until the final message arrives long after anyone cares. Nobody notices, because the platform only ever shows one step at a time.
How to set them properly
Sketch the whole timeline on paper before touching the builder, from trigger to last message, and read it as a customer would: this arrives now, this the following morning, this at the weekend. If the sequence looks relentless written out, it will feel relentless in the inbox.
Use absolute waits wherever the hour matters — sales follow-ups during business hours, order-related messages after a realistic delivery window — and check which time zone the platform is using before you rely on it. Where the tool supports it, send time optimisation can handle the last hop, but it cannot rescue a sequence whose overall shape is wrong. Then test properly: enter the flow yourself with a real address and watch when each message actually lands.
