All posts

We measured a real inbox: 10 senders produced 43% of the noise

Everyone knows their inbox is full of newsletters. Almost nobody knows the shape of it — and the shape is what tells you whether cleanup is worth an afternoon or five minutes.

So we measured ours. Every number below comes from MorningZero‘s production database on 5 August 2026: 8 connected mailboxes, 15,773 mirrored messages, 2,866 of them classified.

Three numbers

93.8% of bulk mail was never opened. Of 704 messages classified as Newsletter or Marketing, 660 were still unread. Notifications were worse: 1,633 messages, 95.7% unread.

Ten senders produced 43% of the noise. Across all 2,337 machine-generated messages (Newsletter, Marketing and Notification combined), from 411 distinct senders, the ten heaviest senders accounted for 1,008 messages — 43.1%.

Under 2% of classified mail needed a reply. 40 of 2,866 messages, as broken down here.

The middle number is the useful one, and it’s the argument against the way most people clean their inbox. If you unsubscribe message by message, you’re paying the same cost for a sender who mailed you twice as for one who mailed you 180 times. Sorting senders by volume first turns “an afternoon of clicking” into ten decisions that remove almost half the traffic.

The caveat, same as always: these are our own working and test mailboxes, heavier on notifications than a typical personal inbox. The concentration pattern is what generalises — a long tail of senders you barely notice, and a short head doing most of the damage. Yours will have a different head.

Why the leak reopens

Three forces, none of them your willpower:

  1. Subscribing is one click; unsubscribing is a form. Every unsubscribe flow is a different page, sometimes a “sorry to see you go” survey. Friction keeps you subscribed.
  2. You notice the sender you opened, not the fifty you didn’t. The cost is spread thin across hundreds of low-grade messages, so it never crosses the threshold where you act.
  3. Re-engagement campaigns are designed to claw back attention you already withdrew. “We miss you”, “your cart is waiting”, “one last chance.”

What Sweeps does about it

Sweeps inverts the friction. Instead of hunting unsubscribe links one message at a time, MorningZero groups your mail into clusters you can act on together:

  • Never-opened senders — the ones whose mail you consistently ignore, bundled so you can cut them in one pass
  • High-volume senders — ranked by how much room they’d free up, which is exactly the 43% head from the measurement above
  • Stale notifications — alerts from tools you’ve stopped using

Each group offers two actions:

  • Unsubscribe, using the sender’s List-Unsubscribe header where they provide one — the standardised mechanism, not a scraped footer link
  • Archive, clearing the backlog those senders already left behind

Because the grouping is done for you, the job becomes a couple of decisions rather than an afternoon.

Unsubscribe vs archive vs delete

ActionWhat it doesUse it when
UnsubscribeStops future mail from the senderYou never want this sender again
ArchiveRemoves existing mail from the inbox, keeps it searchableYou want a clean inbox but might need the history
DeleteRemoves mail permanentlyYou’re certain you’ll never need it

The default combination worth reaching for is unsubscribe plus archive: kill the future flow, tidy the past, destroy nothing. Anything archived stays findable later — including by asking a question about it.

A hygiene routine that survives contact with reality

Sort by volume, not by annoyance. Your instinct is to unsubscribe from whatever irritated you this morning. The measurement says go after the heaviest senders first — that’s where the 43% lives.

Run it monthly, not continuously. Once a month, open Sweeps, look at the grouped clusters, cut what you’ve stopped opening. It takes minutes because the grouping already happened.

Give “not sure yet” subscriptions a holding pen. When you sign up for something you’re ambivalent about, let a natural-language rule file it straight to a Trial label. If you never open the label, the next sweep clears the whole thing.

Keep what you actually read, but batch it. Newsletters you value don’t need to interrupt you on arrival. Let labeling collect them in a Reading bucket you visit deliberately.

Protect the needs-reply view. The entire point of cutting noise is that the one view that matters stays short and trustworthy. If needs-reply is honest, you can ignore everything else without anxiety.

Sweep every account at once. Because the inbox is unified, one pass spans all connected mailboxes — you’re cleaning one stream that happens to draw from three.

What this doesn’t fix

  • Senders who ignore List-Unsubscribe. Some don’t implement it, or implement it badly. Those still need the footer link and a browser tab.
  • Mail you asked for and forgot. A classifier can tell you a sender is bulk; it can’t tell you that you signed up on purpose. That judgement stays yours, which is why Sweeps groups and suggests rather than acting on its own.
  • The transactional middle. Receipts and calendar mail are noise until the day you need one. Those get filed, not cut.

Start with the head, not the tail

If you do one thing after reading this: find your ten heaviest senders and decide about them. In our data that single pass would have removed 1,008 of 2,337 machine-generated messages.

Connect an account and run a sweep — the grouping takes seconds, and the top of that list is usually a surprise.

Frequently asked

What share of newsletters and marketing email never gets opened?

In MorningZero's production data on 5 August 2026, 660 of 704 messages classified as Newsletter or Marketing were still unread — 93.8%. Notifications were worse: 95.7% of 1,633 unread.

Is unsubscribing better than archiving?

They solve different problems. Unsubscribing stops future mail from that sender; archiving clears what already arrived while keeping it searchable. For a sender you'll never want again, do both — unsubscribe treats the cause, archive treats the backlog.

Why does bulk unsubscribe work better than doing it one email at a time?

Because the noise is concentrated. Across 2,337 machine-generated messages from 411 distinct senders, the top 10 senders accounted for 1,008 messages — 43.1%. Cutting ten subscriptions removes nearly half the volume, so grouping senders beats hunting individual unsubscribe links.

Does unsubscribing in MorningZero actually hit the sender's unsubscribe link?

Yes — Sweeps uses the message's List-Unsubscribe header where senders provide one, which is the mechanism mail clients are supposed to use, rather than scraping the footer link out of the HTML.