AI Sales Academy

Spam or inbox? Deliverability in practice

Cold Email 10 min
Business problem

Even a good campaign dies when emails don't reach the inbox. Deliverability isn't 'set once' — it's a system you have to maintain: volume grows too fast, the list has bounces, the content begs for complaints. You don't see the problem because you're 'sending' the emails — it's just that no one reads them.

Proven method

Deliverability in 5 pillars:

  1. DOMAIN REPUTATION: grows with good practices (sending from a warmed domain, replies, low bounce) and collapses in an instant (mass bounce, complaints, spam reports),
  2. LIST: clean, <2% bounces, no bought emails — because that's the main killer,
  3. CONTENT: no spam words, no clickbait, a natural CTA (one question, zero 'guarantee', 'free'),
  4. VOLUME: stable, growing slowly (warmup), a daily limit per domain,
  5. MONITORING: check bounces and replies daily; Gmail has 'Promotions/Spam' — test. Alarm signals: bounce >3%, open rate drops, emails to your own inbox land in spam, complaints grow. Repair: immediately stop sending from the domain, clean the list, check the content (what could have triggered it), resume with a lower volume. AI: analyzes emails for spam triggers before sending, proposes fixes, reports bounces/replies per campaign.
Market example

The 'lead recovery' campaign: you send 40 emails/day from a warmed domain, a verified list (1.2% bounce), no clickbait content. After 3 weeks: 50% opens, 7% replies. Suddenly in week 4 opens drop to 20%. You check: 3.5% bounce — a batch of bad emails got into the list (source: a new data provider). Action: you stop the new batch, remove the emails, return to the old list. After a week opens return to 45%. That's daily deliverability hygiene.

Ready-to-use prompt
You are a deliverability expert.

Context: My campaign: [PRODUCT]. Data: [SENT, BOUNCES, REPLIES, OPENS]. Domain: [DOMAIN]. Volume: [X]/day. List source: [WHERE FROM].

Task: A deliverability audit:
1. Assess the 5 pillars (reputation, list, content, volume, monitoring) — status: good/medium/bad
2. Alarm signals in my data (bounce rate, open trend)
3. Risks: what could burn my domain in this setup
4. A 7-day repair plan (step by step)
5. A daily checklist (5 min) to maintain deliverability

Format: 5 sections.
Model rekomendowany: Claude 3.5 Sonnet / GPT-4o Copy and paste into ChatGPT, Claude or Perplexity
ALTERNATYWNE SCENARIUSZE

Variations & alternative scenarios

Launch version: a checklist before the first campaign.

Scale version: multi-domain monitoring + alerts.

Practical exercise

Calculate your bounces and open rate (or assume numbers). Assess the 5 deliverability pillars in your setup (good/medium/bad). Write down 1 thing to fix this week.

Zapisano w przeglądarce

Steps to take

0/3