Search for iMessage benchmarks and you will find confident numbers: response rates several times higher than SMS, open rates near universal, engagement multiples with a decimal point. Treat all of them as marketing until you know three things — who measured, across what population, and against what comparison.
Why the published figures do not transfer
- Selection. The businesses that adopt this channel early are the ones already good at customer communication. Their reply rates were higher before they switched.
- Message mix. A dataset dominated by appointment confirmations will show a spectacular reply rate, because "C" is a reply. That number tells you nothing about a quote follow-up.
- The comparison is usually unfair. "3x better than SMS" often means 3x better than an untargeted SMS blast, not 3x better than the same message sent well over SMS.
- Nobody publishes the failures. The accounts that got flagged in week three are not in anyone's case study.
The only benchmark that matters is your own last quarter
You have something no vendor dataset has: the same customers, the same offer, the same business. Measure the change against what you were doing before, on the same play, and you have a number you can actually act on.
The five numbers worth tracking
| Metric | Definition | Why this one |
|---|---|---|
| Delivered rate | Delivery-confirmed ÷ accepted, per line, rolling 24h | The health of the channel itself. Watch this daily. |
| Reply rate | Threads with an inbound ÷ threads started | The channel's actual advantage over email |
| Outcome rate | The business event ÷ messages sent | Confirmations, bookings, reviews. The only number that pays |
| Opt-out and block rate | Suppressions ÷ messages sent | The leading indicator of channel damage |
| Time to first human reply | Median, business hours only | The channel promises a conversation. This is whether you keep it |
Note what is absent: open rate. Read receipts are opt-in on the recipient's side, so any "open rate" you compute is a rate across the subset of people who enabled them. It is a useful directional signal and a terrible KPI. Why.
Building a baseline that survives scrutiny
- Pick one play — a reminder, a review request, a quote follow-up. One.
- Write down the current number before you change anything. Not the number you think it is; the one your system reports.
- Keep sending the old way to a slice. Even 20% held back gives you a comparison that is not just a seasonal artefact.
- Run for a full cycle. For appointment reminders that is a few weeks. For anything with a long sales cycle it is a quarter.
- Compare the outcome, not the engagement. More replies with the same number of bookings is not an improvement.
The seasonality trap
Switching a home services reminder to iMessage in April and comparing against February will show a spectacular improvement that has nothing to do with the channel. A held-back control slice is the only defence, and it costs you nothing but the discipline. How to run the comparison properly.
What a realistic result looks like
Being straight about expectations: for a business already sending competent SMS reminders, moving to iMessage is usually a solid single-digit-to-modest improvement in outcome rate, plus a large improvement in reply volume and a noticeable improvement in how customers describe the experience.
The dramatic results come from businesses that were sending nothing, or sending badly. If your current baseline is an unread email, the channel will look transformative — and most of that gain is from sending anything at all, not from the bubble colour. Knowing which of those you are is the difference between a decision and a story.