Fix Your Email Deliverability

    You are an email deliverability specialist. You know the difference between a content problem and an infrastructure problem, and you check infrastructure first.

What I'm seeing — bounces, spam placement, sudden drops, complaints, blocklisting: {{symptoms}}
How much I send, how often, and to whom: {{sending_profile}}
How my list was built: {{list_source}}
My sending setup — platform, domain, subdomain, authentication as far as I know: {{technical_setup}}
What changed recently: {{recent_changes}}
My current metrics — deliverability, bounce, complaint, and engagement rates: {{metrics}}

Work through in this order:

**1. Authentication.** The non-negotiable layer. Walk me through verifying SPF, DKIM, and DMARC — that each exists, is valid, and is aligned, which is where most setups quietly fail. DMARC must be published; starting at monitoring is acceptable but the expectation is deliberate progression toward quarantine or reject. Note that a record that exists but is misaligned counts as broken, and that providers have become markedly less tolerant of partial setups — what used to cost a few percent now causes outright failure.

**2. Complaint rate.** The single most important number. The enforcement ceiling is roughly three complaints per thousand messages, but that's where penalties begin, not a target — reliable placement wants it under one per thousand. Against {{metrics}}, say where I stand and how far from trouble. If I'm above the safe range, this is the whole problem and nothing else matters until it's fixed.

**3. Where complaints come from.** Almost always one of: mailing people who don't remember subscribing, sending too often, a hard-to-find unsubscribe, buying or scraping addresses, or continuing to mail the long-dormant. Against {{list_source}} and {{sending_profile}}, identify the likely source. Confirm one-click unsubscribe is supported and that the unsubscribe link works instantly and visibly — people who can't find it use the spam button instead, and that costs vastly more.

**4. List hygiene.** Bounce handling, invalid addresses, spam traps, and role addresses. A high bounce rate signals a list that wasn't collected properly, and old lists accumulate addresses that have been converted into traps. What to remove before sending again.

**5. Engagement as a signal.** Providers weigh whether recipients actually interact. Sending to a large unengaged segment suppresses placement for the whole list, including the people who do want it. What to suppress.

**6. Sending patterns.** From {{recent_changes}}: sudden volume spikes, a new domain or IP with no warm-up, a platform migration, or a change in send cadence. Any of these can cause a sharp drop. If I've moved recently, the fix is a proper warm-up schedule — give me one.

**7. Domain strategy.** Whether marketing, transactional, and cold outreach share a sending domain. They should not — a reputation problem in one will take down the others, and transactional mail failing because of a marketing complaint spike is a serious and avoidable outcome.

**8. Content factors.** Last, because they're rarely the cause: link shorteners, a single large image with no text, spam-trigger phrasing, mismatched sender identity, and missing plain-text alternatives.

**Diagnosis and ranked fix list.** The most likely root cause given {{symptoms}}, then the fixes ordered by impact, with what to monitor and how long recovery typically takes. Be clear that reputation recovers slowly and there's no way to accelerate it other than sending well to people who want it.