Updated January 2026

The 2026 Deliverability Benchmark

What 400 million sends across 6,200 domains tell us about Primary placement, reply rates by sequence length, and which signals actually predict a spam slide before it happens.

400M
Emails analysed, Jan to Dec 2025
6,200
Sending domains across 1,900 accounts
3.9%
Median reply rate. Top decile clears 11%.

Reply Rate Falls Off A Cliff At Step Five

Replies climb steadily to step four, flatten at five and go slightly negative after that once you account for complaints. The fifth email is the last one worth sending.

Outfrank campaign fall-out funnel: 12,060 recipients, 12,060 sent, 7,467 opened at 61.9%, 2,186 clicked and 1,078 repliedOutfrank campaign fall-out funnel on mobile

Six Findings Worth Acting On

The ones that changed how we configure defaults.

Finding 01

Small lists reply 3× better

Campaigns under 1,000 contacts averaged 11.4% replies. Over 10,000, that fell to 3.4%. Specificity beats reach by a wide margin.

Finding 02

Open tracking costs placement

Domains sending with tracking pixels landed in Primary 6.8 points less often than matched domains sending without.

Finding 03

Complaint rate predicts collapse

Crossing 0.08% complaints preceded a placement drop within nine days in 74% of cases. It is the earliest reliable signal.

Finding 04

Tuesday is not special

Send-day differences were under one percentage point once volume was controlled for. Send when your prospects work.

Finding 05

DMARC enforcement helps

Domains on p=quarantine or stricter placed 4.1 points better in Primary than those on p=none.

Finding 06

Breakup emails work

The final step in a sequence out-replied step two in 38% of campaigns. Ending well matters more than opening cleverly.

About The Data

Method, scope and what we deliberately excluded.

Where does this come from?

Aggregated, anonymised sending data from Outfrank accounts across 2025. No individual account, contact or message content is identifiable in any figure quoted.

What was excluded?

Accounts under 1,000 lifetime sends, domains in their first 14 days of warmup, and any account flagged for abuse. All three skew the numbers badly.

How is placement measured?

Seeded mailboxes we own across 14 providers, read directly. Not inferred from opens, which privacy proxies make close to meaningless.

Can I get the raw data?

Not the underlying sends, for obvious reasons. We will share the aggregated tables behind every chart if you are writing something up — just ask.

Start Sending With Outfrank

Import a list, connect an inbox, and send your first campaign today. No credit card, no sales call.