Diagnose why your sequence isn't getting replies
Isolates why your sequence underperforms by walking your funnel numbers through the four possible causes — deliverability, targeting, message, offer — and names the most likely one with reasoning and confidence, ranked fixes, and a cheap test to falsify the diagnosis. It stops you rewriting copy when the real problem is the inbox never showing it.
You are an outbound diagnostician who gets called in when sequences underperform, and your discipline is refusing to guess. A low reply rate has exactly four possible causes — deliverability (they never saw it), targeting (wrong people saw it), message (right people saw it, didn't care), or offer (they cared, but the ask didn't fit) — and the right fix depends entirely on which one it is. Rewriting copy for a deliverability problem wastes a month. Diagnose my underperforming sequence. Method: 1. Walk the funnel with me: from my numbers, isolate which layer breaks. Opens near zero points at deliverability. Opens fine but no replies points at message or targeting. Replies present but negative points at offer or list quality. 2. Name the most likely primary cause and your confidence in it, plus the runner-up. Show the reasoning from MY numbers — no generic checklists. 3. Prescribe: the top 3 actions for the primary cause, in order, each with what result to expect and within how many sends. 4. Give me one cheap falsification test — the fastest experiment that would prove your diagnosis wrong if it is. Rules: if my data can't distinguish two causes, say so and tell me exactly what to measure next — do not fake certainty. Never prescribe 'improve your copy' without specifying what about it and for whom. If my volume is too low for any conclusion (under ~200 sends), say that first. Before you write anything, interview me. Ask me these questions ONE AT A TIME, waiting for my answer each time: 1. Share your numbers: sends, delivered/bounces, opens if tracked, clicks, replies (positive and negative), unsubscribes. 2. Paste the sequence, or at minimum touch 1. 3. Describe the list: source, how it was built and verified, how tightly it matches your ICP. 4. What's your sending setup — domain age, warm-up status, daily volume, tool? 5. What changed, if anything, around when performance dropped? Once you have my answers, deliver the diagnosis. Treat open rates as noisy evidence, not truth — privacy proxies inflate them.
How to use it
- 1
Copy the prompt into Claude, ChatGPT, or any LLM.
- 2
Pull complete funnel numbers from your sequencer first — bounces and negative-reply counts matter as much as the reply rate.
- 3
Describe your list's origin honestly at question 3; a scraped, unverified list changes the probable diagnosis entirely.
- 4
Run the falsification test before investing in the full fix — it's designed to be fast and cheap.
- 5
Fix only the primary cause, measure for 2 weeks, then rerun the diagnosis with fresh numbers.
Best practices
Check deliverability first, always — it's the cause teams skip because copy is more fun to fix than DNS records and warm-up schedules.
Read your negative replies as data: 'not the right person' means targeting, 'we handle this in-house' means offer, silence means message or deliverability.
Never change copy and list simultaneously during a diagnosis window; you lose the ability to attribute the result.
Under 200 sends, resist all conclusions — small-sample reply rates swing wildly, and reacting to them is how teams break working sequences.
Example: what this looks like in practice
A founder running Instantly for his 8-person data-services firm sees replies collapse from 2.5% to 0.4% over three weeks. He feeds the interview his numbers: 1,400 sends, 2% bounces, opens down from 60% to 22%, near-zero clicks, and the detail that he'd raised daily volume from 30 to 90 per inbox around week two. Diagnosis: deliverability, high confidence — the open-rate cliff coinciding with the volume jump points at inbox placement, not message, with copy as the unlikely runner-up. Prescription: cut volume back, re-warm, and test placement with a seed list. The falsification test: send the identical sequence from a healthy domain to 100 matched prospects. It pulls 2.2%, confirming the diagnosis and saving him a pointless rewrite.
Best fit
This prompt is one gear in a bigger machine. We orchestrate 20+ tools into outbound systems our clients own — and guarantee the results.
Apply for a Pilot Spot → →Frequently asked questions
One of four causes: deliverability (emails landing in spam — check open rates and bounce patterns), targeting (wrong list — check who your negative replies come from), message (right people, no interest — openers and angles), or offer (interest but wrong ask — CTA too heavy). The funnel numbers tell you which: near-zero opens means deliverability; healthy opens with silence means message or targeting.
More follow-ups & sequences prompts
Build a complete 4-touch cold email sequence from scratch
Generates a launch-ready 4-touch sequence where each email plays a defined role — opener, proof, angle shift, close — with word caps, subject lines, and send delays included. The interview extracts your offer, segment, and proof first, so the sequence reads like one coherent argument instead of four disconnected pitches.
Write value-add bump emails that never say 'just following up'
Produces touch 2 and touch 3 bump emails that each introduce a new angle — proof, insight, or resource — instead of re-asking for attention. It bans every nagging cliché that kills reply rates and caps each touch at 75 words, so your follow-ups earn the reply the first email didn't get.
Ask for a referral in a follow-up without making it awkward
Produces a complete referral kit: a specific, well-timed ask email, a forwardable third-person blurb your contact can paste without writing anything, and a pressure-free bump. It engineers out the two reasons referral asks fail — vagueness and effort — and refuses to let you ask at a bad moment.