Deliverability Case Study: "I Know Better"
A retro soul, 60s-R&B song told entirely from inside the head of the villain of its own story: a CEO who's certain his gut beats any dashboard, and who reads every deliverability warning as a reason to send more, not less, because more volume means more revenue. There's no persona bit here and no punchline; the song is a straight character study of Pride — one of email marketing's seven deadly sends — where short-term send volume gets priced as guaranteed revenue and the deliverability cost gets priced as somebody else's problem, right up until the mailbox providers stop asking politely and start enforcing.
Here is the technical breakdown of the deliverability concepts detailed in the song:
Verse 1 & Pre-Chorus: Refusing the Three Signals That Actually Decide Placement
"He talked about engagement / He talked about complaint rates / He talked about reputation / I said, 'You're looking at the wrong data'"
- The Deliverability Context: Engagement, complaint rate, and sender reputation are not three of many metrics — they're the three inputs mailbox provider filtering models weight most heavily when deciding inbox versus spam folder. None of them are fully visible to the sender in real time: engagement is inferred from opens, clicks, and replies; complaint rate maps to the spam-report data mailbox providers collect internally; and reputation is a rolling internal score a sender only sees indirectly, through worsening placement, now that Google Postmaster Tools' v2 migration has dropped the old domain and IP reputation tiers. Calling all three "the wrong data" isn't a rebuttal — it's clearing the last obstacle standing between him and sending at whatever volume he wants.
- The Volume Claim: "If I want to send a million, I'm gonna send a million" is the line the whole song hangs on: volume as a decision made by willpower and revenue math alone, with deliverability capacity treated as irrelevant. In reality, inbox providers evaluate volume relative to established sending history — a sudden jump that isn't backed by a warmup ramp reads as anomalous regardless of intent, and gets throttled or filtered on that basis alone. The "million" isn't a number the sender gets to just decide; it's a number the receiving side has to be willing to accept first.
Verse 2: Excuses That Don't Survive Contact With the Mechanism
"They said, 'Your opens are falling' / I said, 'Apple changed the game'"
- The Excuse That Proves He Doesn't Understand the Mechanism: This isn't a deflection with a kernel of truth in it — it's backwards, and it's the first hard evidence in the song that the confidence was never backed by real technical knowledge. Apple's Mail Privacy Protection (MPP), shipped with iOS 15, pre-fetches remote content for virtually every Apple Mail user the moment a message reaches the mailbox, which pushes reported open rates up, not down, regardless of whether a human ever reads the message. MPP cannot produce a falling open rate under any scenario — it only inflates. So if opens are genuinely falling, that has nothing to do with Apple and everything to do with mail not reaching the inbox in the first place: the exact placement problem the verse names two lines later. Blaming "Apple changed the game" isn't a wrong guess about a real phenomenon, it's citing a mechanism that runs in the opposite direction of the symptom he's explaining — which means the CEO doesn't actually know what MPP does, he's just reaching for the nearest deliverability-sounding word.
- The Threshold Being Ignored: "The complaints are rising / I said, 'People complain sometimes'" is true in isolation and irrelevant to the actual rule. Gmail treats a sustained complaint rate above roughly 0.10% as a warning signal and above roughly 0.30% as grounds for active filtering or blocking — the threshold doesn't care whether any individual complaint feels routine.
- Where Denial Meets the SMTP Layer: "Then the inbox started shrinking / And the blocks began to grow" is the moment internal explanations stop mattering. This is mailbox providers moving from soft signals to hard enforcement — bulk-foldering mail server-side or issuing outright 5xx rejections at the MTA. No spreadsheet argument changes what's happening on the wire.
Bridge: Four Standard Fixes, Rejected in the Same Breath
"Maybe stop mailing people who haven't engaged in years." / I said, "That's revenue." / ... / "Maybe listen." / I said, "That's not how we do things here."
- The Deliverability Context: Every suggestion in this bridge is a textbook remedy, not a guess: suppressing chronically unengaged subscribers is a sunset policy (list hygiene); suppressing people who've already complained keeps complaint rate under threshold; reducing volume matches send rate to actual list quality and warmup pace. Each one gets rejected with the same one-line math: "that's revenue" — the CEO pricing the fix's cost (fewer sends, a smaller list, a slower ramp) as certain lost income right now, against a reputational payoff he treats as hypothetical. The bridge doesn't show a sender who's missing information; it shows one holding all four standard fixes and rejecting every one because each would mean sending less, and sending less isn't compatible with a strategy built on volume driving revenue. "Maybe listen" gets the final, flattest refusal — "that's not how we do things here" — because by that point the CEO isn't weighing evidence anymore, he's just restating the plan.
Verse 3 & Pre-Chorus: Why Reputation Breaks Asymmetrically
"Every warning light was flashing / But I wasn't worried yet"
- The Asymmetry: "Because sometimes reputation / Takes a month to build... / And one stupid week." This is an accurate read of how reputation models actually behave. Gmail, Microsoft, and Yahoo build sender reputation from rolling, trailing engagement and complaint history, which is exactly why reaching a "High" tier takes sustained weeks of clean sending — but one anomalous, high-volume, low-quality batch can crater that trailing average in days. Reputation is slow to earn and fast to lose because the underlying math is an average, not a single score.
- The Payoff: "He didn't say, 'I told you.' / He didn't have to speak." Nobody needs to argue anymore — the inbox metrics are now saying what four verses of confidence couldn't argue away. Every warning dismissed as "the wrong data" in verse one is now sitting in the numbers as fact.
Bridge 2 & Final Chorus: The Sender Doesn't Get the Final Vote
"Pride says, 'Send it.' / Reputation says, 'Don't.' / ... / The mailbox says, 'No.'" and "When you think you know everything / The filters know you well"
- The Deliverability Context: This is the song's thesis stated outright. Confidence and internal arguments are not inputs the mailbox provider's filtering model consults — only behavior is. "The filters know you well" holds up technically: Gmail's filtering weighs a sender's aggregate history and recipient behavior over time as heavily as any single message's content, so the record being judged at send time is the sender's own, and it's a record he's been actively building against himself for verses.
There's no technical fix offered in the outro, and that's the point — nothing here was a missing DNS record or a misconfigured header. The failure was behavioral the whole way through, and the decision that caused it was made before the campaign was ever scheduled.
If "just send more" has ever been the answer to a revenue target, this song is the fast-forward version of what that math misses. Here's how to price the deliverability side of that equation before a block prices it for you.
Do the Revenue Math All the Way Through
More volume only means more revenue if the mail actually lands — past a certain point, it means the opposite.
- Treat send volume as bounded by list quality and reputation, not by ambition. A warmup ramp exists because "we can technically send it" and "the inbox will accept it" are two different questions, and Pride only ever answers the first one.
- Price the other side of the ledger. Suppressing unengaged users, honoring complaint suppression, and slowing volume all cost send count today — but mail that's blocked or foldered generates zero revenue no matter how large the list was, and rebuilding reputation after a placement collapse takes far longer than the send that caused it.
- Don't let a revenue target override a deliverability constraint. If hitting a number requires ignoring engagement, complaint rate, or reputation signals, the actual constraint is inbox capacity, not sales effort — no amount of confidence changes what the receiving side will accept.
Trust the Three Core Signals Over Your Gut
Engagement, complaint rate, and sender reputation are the model — not competing opinions about the model.
- Check Postmaster Tools v2's compliance and deliverability-audit data weekly. Google dropped the old domain/IP reputation tiers in mid-2026, but v2 now gives a direct compliance readout and a summary verdict on whether your mail is wanted or driving engagement versus complaints — arguably a clearer signal than the old tiers were. Round it out with Spamhaus/blocklist status and Microsoft SNDS for the Outlook/Hotmail side.
- Treat a metric moving the wrong way as information, not noise. If it contradicts your instinct, investigate the metric — your gut has no visibility into a filtering decision happening inside someone else's mail server.
- Don't wait for a hard block to believe a soft warning. Blocks are the lagging confirmation of a reputation problem that engagement and complaint data already flagged.
Know Which Metrics Are Actually Broken vs. Which You're Choosing to Ignore
- Open rate is inflated, not just noisy, by Apple's Mail Privacy Protection (MPP). It pre-fetches content on delivery, pushing reported opens up regardless of whether a human ever reads the message — so a falling open rate despite that inflation is a placement red flag worth investigating immediately, not a metric to explain away.
- Delivered isn't the same as inboxed. A high delivery rate only means the receiving server accepted the message at SMTP time — it says nothing about whether that message landed in the inbox or the spam folder.
Treat Complaint Rate as a Hard Limit, Not a Vibe
- Enroll in every Feedback Loop (FBL) your mailbox providers offer and auto-suppress anyone who complains — a complaint is an unsubscribe with extra urgency.
- Keep complaint rate under Gmail's ~0.10% caution line. Above ~0.30% is documented blocking territory, not a range where "people complain sometimes" is a defense.
- Suppressing complainers isn't lost revenue — it's the cheapest reputation insurance available, and refusing it is what turns a soft complaint signal into a hard block later.
Run a Real Sunset Policy
- Suppress or re-permission subscribers unengaged for 90–120 days before mailbox providers quietly stop trusting their continued presence on your list — and before an abandoned address gets recycled into a spam trap.
- Segment by engagement before scaling volume up, rather than mailing the full list at once and hoping the good addresses carry the bad ones.
- Ramp volume increases gradually, matching warmup pace — a sudden jump to your ceiling volume looks anomalous to a filter no matter how confident you feel about it.
Conclusion
Reputation compounds slowly and breaks fast — a CEO chasing more volume for more revenue on confidence alone is trading one extra week of sending for a month or more of rebuilding the trust that got him inbox access in the first place. Price the deliverability cost into the volume decision before a block prices it for you.
Your Reputation & Pride Checklist:
- Price the deliverability cost of any volume increase before you send, not after the block.
- Check Postmaster Tools v2's compliance and deliverability-audit verdict weekly — old IP/domain reputation tiers are gone since mid-2026, but v2 now tells you directly if your mail is wanted.
- Don't trust a falling open rate post-Apple MPP — MPP inflates opens, so a drop is a placement red flag, not noise.
- Enroll in FBLs and auto-suppress complainers immediately.
- Keep complaint rate under 0.10%; treat 0.30% as a hard block threshold, not a shrug.
- Run a 90–120 day sunset policy and ramp volume increases gradually instead of jumping to full send.
Educational content. Email deliverability evolves rapidly. Platform rules (Gmail, Yahoo, etc.), engagement signals, and ESP behaviours change frequently, and real-world issues often involve conflicting signals, data quality problems, and failure modes that general best practices can’t anticipate. Content on this site is provided for informational purposes only and does not replace a thorough analysis by a qualified deliverability professional.
Terms of Use