New
What the Algorithm Knows
An original song by Inbox Senders Club
AlbumClick ThroughTrack 6Three A.M. after a blocklisting. Marcus Deliverino finally reads the documentation and learns what he should have known before his first blast: ISPs don't judge tabs — they judge engagement. Opens, clicks, replies, spam reports. The tab placement was never the enemy. The signal was.
3:49
Duration
Verse 1
Domain's dead. I'm reading through the docs.
Three A.M. — alone with nothing but the clock.
The answer cut right through the noise:
They watch what your subscribers do — opens, clicks, destroys
Chorus
What the algorithm knows —
The tab you're in just follows where engagement goes
Opens mean you earned it
Complaints mean you burned it
That's what the algorithm knows
Verse 2
They track if someone opened what I sent
If they clicked, replied — or if they went
Straight to spam without a second read
That's the signal that decides if I succeed
Chorus
What the algorithm knows —
The tab you're in just follows where engagement goes
Opens mean you earned it
Complaints mean you burned it
That's what the algorithm knows
Bridge
I declared war on the Promotions tab
Now I see I missed the point by half —
The inbox isn't placed — it's earned
One lesson learned, one domain burned
Chorus
What the algorithm knows —
The tab you're in just follows where engagement goes
Opens mean you earned it
Complaints mean you burned it
That's what the algorithm knows
Outro
From here — engagement only
No more tabs, no more tricks
Just the work
Just the signal
Just the trust
Deliverability Case Study: "What the Algorithm Knows"
Track 06 of Click Through is the album's turning point — the moment Marcus Deliverino stops fighting and starts listening. His domain is gone, his campaigns are dead, and at three in the morning he finally reads what every sender should have read before their first blast. The lesson is simple, uncomfortable, and career-changing: inbox placement isn't something you engineer around a tab. It is the direct output of how your subscribers interact with every email you send.
Here is the technical breakdown of the engagement signals and filtering mechanics detailed in the song:
Verse 1: Per-User Engagement Is the Only Signal That Matters
"They watch what your subscribers do — opens, clicks, destroys"
- The Deliverability Context: Modern mailbox providers — Gmail, Yahoo, Outlook — do not apply a single global filter to your messages. They build an individual engagement model per recipient. Each subscriber has a personal history with your domain: how often they open, whether they click, whether they move your email to trash without reading it. That per-user record is the core input to tab placement, filtering, and ultimately inbox delivery.
- "Destroys" as a Signal: The lyric is blunter than it sounds. Deleting an email without opening it is a tracked negative signal. Done at scale, it tells the algorithm that your mail is unwanted noise. ISPs call this a passive complaint — no spam report button pressed, but the behavioral data reads the same way.
Verse 2: The Full Signal Stack — Positive and Negative
"They track if someone opened what I sent / If they clicked, replied — or if they went / Straight to spam without a second read"
- The Deliverability Context: This verse correctly enumerates the complete engagement signal stack. Positive signals — opens, clicks, replies, moving mail out of spam, adding the sender to contacts — accumulate into reputation credit. Negative signals — spam reports, deletions without opening, never interacting across multiple campaigns — drain it. Your domain reputation at Gmail Postmaster Tools is essentially a weighted aggregate of these signals across your entire subscriber base.
- "Straight to spam without a second read": This is the scariest metric most senders never see. Unlike an active spam report (which triggers Feedback Loop data your ESP can capture), passive deletions and inbox-to-spam moves are invisible to the sender but fully visible to the ISP. The only way to reduce them is to send mail people actually want to read.
Chorus: Tab Placement as a Lagging Indicator
"The tab you're in just follows where engagement goes"
- The Deliverability Context: Gmail's tabbed interface (Primary / Promotions / Social / Updates) is commonly misunderstood as the cause of low engagement. It is the effect. Gmail classifies messages using a combination of content signals and — critically — per-user behavior. If a subscriber consistently opens your email in Promotions, clicks through, and replies, Gmail will, over time, begin placing your messages in Primary for that user. There is no header trick or subject line hack that bypasses this. The tab you're in is a readout of your engagement health, not an obstacle to it.
Bridge: The Wrong Battle — Admitting the Mistake
"I declared war on the Promotions tab / Now I see I missed the point by half"
- The Deliverability Context: The bridge subverts Track 02 of the album directly. "Promotions Is the Enemy" was Marcus's opening thesis — and it is the single most common deliverability misconception in the industry. Senders spend engineering hours testing subject line keywords, email preview text, and send-time hacks designed to route around the Promotions tab. Every hour spent on those tactics is an hour not spent on authentication, list hygiene, segmentation, and relevance — the work that actually compounds into earned inbox placement.
- "One domain burned": The cost of fighting the wrong battle is measured in domain reputation. A burned domain cannot be warmed back up quickly. Many senders abandon it entirely and start over on a new subdomain, carrying the same bad habits with them.
Most senders who struggle with inbox placement are fighting the right war on the wrong front. They audit subject lines, test send times, and debate tab classification — while ignoring the signal stack that drives every one of those outcomes. Engagement is not a metric. It is the mechanism.
The Signals ISPs Actually Measure
Mailbox providers build per-user engagement models from behavioral data, not just content scans. Understanding what goes into that model is the first step to improving it.
- Positive signals — opens, clicks, replies, forwarding, moving mail out of spam, adding your address to contacts. Each one increases your reputation credit with that specific recipient.
- Negative signals — spam reports, deleting without opening, inbox-to-spam moves, never interacting across multiple campaigns. These drain reputation credit and are partially invisible: your ESP's Feedback Loop (FBL) captures active spam reports, but passive deletions and behavioral churn are only visible to the ISP.
- Complaint rate thresholds: Gmail issues warnings above 0.10% and takes enforcement action above 0.30%. Yahoo applies similar thresholds. These are aggregate rates — if your list is large enough, even a handful of disengaged segments can push you over.
Tab Placement Is a Lagging Indicator, Not a Target
Gmail's tabbed inbox classifies messages per-user based on engagement history. Treating it as an obstacle to hack is the wrong frame entirely.
- The tab follows the signal. If subscribers consistently engage with your mail in Promotions, Gmail will migrate it toward Primary for those users over time. The inverse is also true: tab hacks that trick a message into Primary don't improve the underlying engagement, so the algorithm corrects course within a few sends.
- Stop optimizing for classification. Every engineering hour spent on tab routing is an hour not spent on authentication, segmentation, or content relevance — the work that produces durable inbox placement.
- Measure inbox placement rate directly. Tools like GlockApps or Validity/250ok use seed addresses at major providers to tell you where mail actually lands. Use them instead of inferring placement from open rates, which Apple's Mail Privacy Protection (MPP) has made unreliable since iOS 15.
Passive Complaints Are Invisible and Expensive
Active spam reports are only part of the negative signal picture. Passive disengagement — deletions without reading, inbox-to-spam reclassification — accumulates silently until it surfaces as a reputation drop.
- Monitor Gmail Postmaster Tools. The Domain Reputation dashboard tracks your reputation tier (Bad / Low / Medium / High) and spam rate directly from Google's data. A drop from High to Medium is an early warning that passive signals are accumulating.
- Don't wait for blocklisting. By the time your domain appears on Spamhaus or your IP is throttled, months of negative signal have already compounded. Reputation damage is slow to build and slow to repair — catch it early.
- Segment by engagement before every send. Keep your active segment (opened or clicked in the last 90 days) separate from your re-engagement segment. Sending the same content to both collapses your aggregate engagement metrics and hides the damage done by the inactive tail.
Building Engagement-Led Sender Reputation
Earned inbox placement is the compound result of sending relevant mail to people who chose to receive it, consistently, over time. There are no shortcuts, but there is a repeatable process.
- Send to people who asked. Explicit opt-in — confirmed or double-confirmed — produces lists with higher baseline engagement and lower complaint rates than any growth hack.
- Respect sending cadence. Frequency mismatches (promising weekly, sending daily) are a primary driver of unsubscribes and spam reports. Set expectations at opt-in and honor them.
- Implement one-click unsubscribe (RFC 8058). Required by Gmail and Yahoo since February 2024 for bulk senders. A frictionless unsubscribe reduces spam reports — subscribers who can't find the unsubscribe link reach for the spam button instead.
- Sunset disengaged subscribers. Suppress anyone inactive at 90–120 days after a re-engagement campaign. Inactive addresses accumulate into recycled spam traps and drag complaint rates up across your active sends.
Conclusion
The algorithm knows because it watches. Every send either builds or erodes the per-user engagement record that determines where your next message lands. Tab placement, spam filtering, and inbox delivery are all outputs of the same input: whether the people on your list actually want what you're sending them.
Your Engagement Signal Checklist:- Monitor Gmail Postmaster Tools domain reputation weekly — catch drops before they become blocklisting.
- Keep complaint rate below 0.10% at all times; never approach 0.30%.
- Segment active (90-day) from inactive subscribers before every bulk send.
- Implement one-click unsubscribe (RFC 8058) on all bulk sends.
- Suppress chronically inactive subscribers after a failed re-engagement campaign.
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
Covered Topics
Glossary Terms
Complaint RateRe-engagementInbox Placement RateBulkingDomain ReputationIP ReputationBlocklistSpam TrapGreylistingSender Score
View full glossary →Deliverability is a moving target. This content reflects our best understanding at time of writing — but RFCs get updated, ISP policies shift, and best practices evolve. Spot an error or outdated info? Let us know and we'll fix it.
Credits
Song: “What the Algorithm Knows” — Inbox Senders Club
Songwriter: Inbox Senders Club
Producer: Inbox Senders Club
© Inbox Senders Club. All rights reserved.