A 90%+ open rate on SMS gets quoted like a finish line. Every deck, every case study, every vendor pitch leads with that number, because it's high and it's flattering. It also says almost nothing about whether the message did its job.
Open rate tells you the shopper looked. It doesn't tell you what happened after that.
The metric nobody leads with
Response rate is the number that predicts recovered revenue, and most decks never put it next to open rate. A shopper who reads a text and texts back a real question, "does this come in blue," "why is shipping so expensive," "can I still get free returns" is doing something an open never captures. She's telling you the exact thing standing between her and checkout.
The reply is the whole ballgame. A reply that gets an answer can turn into a sale. A reply that lands in a queue nobody's watching, or gets a generic "thanks for reaching out!" back, produces the same result as no reply at all. The shopper asked a question and never got one answered.
Most brands can't tell you their response rate, because most tools weren't built to report it. Open rate and click rate are easy to pull from a blast platform's dashboard. Response rate requires a system built around reading replies, not just counting who opened. Ask a store how many carts their SMS flow recovered and they can pull the number in seconds. Ask how many shoppers replied with a question, and how many of those questions got answered, and most teams shrug.
Why blast tools hit a ceiling
Klaviyo runs flows. Postscript and Attentive run blast sends at scale, both templated, both one-directional by design. None of that is a criticism of the tools; they were built to do what they do, and they do it well. The ceiling isn't a bug in the product. It's a consequence of the architecture underneath it.
A platform built to send doesn't need infrastructure for reading what comes back. Sending at volume is the whole job, so the reply pipe is an afterthought, if it exists at all. A shopper's answer might land in a shared inbox, or nowhere a human ever checks. Either way, the next scheduled message fires whether or not that reply got handled.
More volume doesn't raise the ceiling. It just means more replies stacking up in a channel nobody built to process them. A brand that doubles its SMS send volume doesn't double its recovered revenue, because the bottleneck was never send volume. It was always the reply.
Agencies see this pattern across every account they run. A client asks for a bigger SMS list, a higher send frequency, a more aggressive discount ladder, all reasonable requests aimed at the wrong lever. Volume was never the constraint. The constraint was what happened the moment a shopper wrote something back.
What moves the number
A system built to read a reply and act on it doesn't carry that ceiling, because response volume and resolution capacity scale together instead of pulling apart. Send more messages, get more replies, and the system built to answer them keeps up, because answering was the job it was built for from the start.
Real two-way conversational SMS runs response rates of 45-60%. That number doesn't exist for one-way blasts, because a one-way channel isn't built to be replied to in a way that changes what happens next. The architecture decides the ceiling before a single message goes out. Templated blast tools cap out around the volume they can send. Conversational systems cap out around the volume they can resolve, and that number moves with better handling, not just bigger lists.
Put two brands side by side, both sending the same SMS volume to the same size list. One runs a templated blast. The other runs a system built to read and answer replies. Same send count, same list size, different revenue outcome, because the variable that predicts revenue was never send volume. It was what happened after the shopper wrote back.
The gap this leaves behind
A flow already running gives a brand the sense that cart recovery is handled. The box is checked, the sequence fires, the open rate looks healthy on a dashboard. None of that measures whether a shopper's real question got answered, and that gap is where revenue leaks out of a system that otherwise looks fine.
This is the same mechanism from the flow-versus-conversation gap: a scheduled message can't read what comes back, so it can't act on it either. Response rate is where that gap becomes visible, once someone bothers to measure it instead of leaning on open rate as a stand-in for success.
Where this leaves the stack
None of this argues for replacing Klaviyo or Postscript. Blast infrastructure still has a job: reaching a list at scale, on a predictable schedule, at a low cost per send. The ceiling matters at the specific moment a shopper replies with a real question, and that's the moment a blast tool was never built to handle.
Running alongside the existing stack, a conversational layer picks up that moment, the one where volume stops being the limiting factor and resolution starts being the whole game.
Curious what your real response rate looks like?
Run a Revenue Proof →