Why “reply rate” misses the point

A percentage without context is a vanity metric. What I pair it with before I change a single word.

“Our reply rate is 9%” sounds scientific. Without context, it’s a mood ring — pretty, not predictive.

Reply rate is a ratio: replies divided by sends. It hides who you messaged, why they replied, and whether any of it moved work forward.

What reply rate alone can’t tell you

  • Quality of replies. “Not interested” counts the same as “yes, Tuesday works” in a naive tally.
  • List quality. A high rate on a tiny, hand-picked list isn’t comparable to a scraped list with the same percentage.
  • Stage of the funnel. Early experiments should teach you relevance — not close deals.

What I track alongside it

Qualified conversations started — however you define “qualified” for your stage. Even a rough tag beats a raw percentage.

Time-to-reply (rough buckets: same day, week, never). Slow silence means something different from fast ignore.

Hypothesis per batch — so I know what I changed when the number moves.

The trap of optimizing the wrong numerator

If I chase reply rate by widening who I message, I’ll get more noise replies — or politeness — and call it success. I’d rather have a lower rate on the right people than a trophy metric on the wrong list.

When reply rate is still useful

Inside a stable ICP and message structure, week over week, it’s a smoke alarm. A sudden drop means something broke — targeting, tone, or deliverability — not that the universe hates you.

Measure replies — but pair the number with context, or you’re polishing a scoreboard for a game you’re not actually playing.