Fix sheet: Cold reply rate is below baseline

This sheet appended when your report flagged the `reply-rate` weakness.

What the numbers actually mean

Cold LinkedIn reply rate industry baseline is 5-12%. Below 5%, the sequence is leaving most of the pipeline on the table; below 2% the sequence is broken at touch 1 AND touch 2. Above 12% means either strong personalization is working or the audience is unusually warm.

The thing most teams miss: the bottleneck is almost always touch 2 and 3, not touch 1. Touch 1 gets people to accept; touches 2-4 are where the actual reply happens, and most teams write a great touch 1 and then template the rest.

The 3 things to fix this week

  1. Audit the gap between touch 1 and touch 2. Read your touch 2 message. If it starts with "thanks for connecting!", you have just wasted the warmest 30 seconds of attention you will ever get. Touch 2 should open with ONE specific question the prospect has an opinion on, framed in a sentence.
  1. Add touch 3 if it does not exist. 70% of B2B LinkedIn replies arrive between touches 2 and 4. Skipping touch 3 because "they did not reply to touch 2" leaves a majority of the pipeline on the table. Use the touch 3 template in the playbook — it acknowledges silence without apologising for it.
  1. Move to email by touch 4. Same conversation, different channel. Senior B2B targets read email more consistently than LinkedIn DMs once their inbox is crowded. The touch 4 template in the playbook is built for this — it references the LinkedIn exchange explicitly so the channel switch does not feel cold.

What "fixed" looks like

Combined sequence reply rate climbs to 8-15%. Distribution should spread across touches 2/3/4/5 — if all replies are on touch 2, the back-half is weak; if all replies are on touch 5, the early-half is weak. Even spread = healthy sequence.

How BIGM solves it

We generate the full 5-touch sequence from a single ICP brief — touches 2-4 get AI-rewritten per-prospect openers, and we re-write underperforming variants live based on per-tenant reply data. Replies compound on the back half of the sequence as designed, not on touch 1 luck.

Score your account against the full rubric.
60 seconds. 8 questions. 0-100 score + ban-risk band.
Take the 60-sec check →