Cold Email Deliverability Benchmarks for 2026: What to Track and What Good Looks Like

Why most deliverability benchmarks mislead

Search for "cold email benchmarks" and you will find confident tables of inbox placement by TLD, by provider and by day of warmup. Treat them with suspicion. Three problems come up again and again:

• Inbox placement can't be read from delivery logs. When a receiving server accepts a message it returns a 250 response. That means "accepted", not "inbox". The server decides on the folder afterwards and does not tell the sender. The only ways to see the folder are seed tests (addresses you control at each provider) or data from the recipient's side. Any benchmark that claims to measure inbox placement "through SMTP response codes" is measuring acceptance, not placement.
• Open rates are inflated. Apple Mail Privacy Protection loads tracking pixels automatically, and some security scanners do the same, so an "open" often isn't a person. Tracking pixels can also hurt cold email deliverability. Many cold senders now turn open tracking off.
• Marketing and cold email get mixed together. Newsletter benchmarks come from opted-in lists. Cold email goes to people who never signed up, so complaint and reply rates behave differently.

That leaves a short list of numbers that are both measurable and backed by someone who publishes them. Those are below.

The benchmarks worth tracking

The sections below explain each one, what the source actually says, and what to do when you are outside the range.

Spam complaint rate: under 0.1%, never 0.3%

This is the one benchmark that the mailbox providers publish themselves. Google's email sender guidelines require all senders to keep the spam rate reported in Postmaster Tools below 0.3%, and Google's sender guidelines FAQ says to keep it below 0.1% and prevent it from ever reaching 0.3%. The same FAQ says bulk senders above 0.3% are not eligible for mitigation from Google until they stay below it for 7 consecutive days. Yahoo's sender best practices also set the line at 0.3%.

0.1% is one complaint per 1,000 emails. On cold email that is a tight budget, and it is why targeting matters more than copy tricks. A complaint is the strongest negative signal a recipient can send.

How to measure it: Postmaster Tools shows Gmail's spam rate for your domain once you send enough daily volume to personal Gmail accounts; at cold email volumes per domain the data is often sparse. Most sending tools record "marked as spam" feedback where the receiving provider reports it. Watch both, and treat any complaint spike on a campaign as a reason to stop that campaign.

Bounce rate: under 2%

No mailbox provider publishes a bounce threshold, so this benchmark comes from sending platforms:

• Instantly's Cold Email Benchmark Report 2026 advises keeping bounce rates below 2%, "ideally much lower".
• Mailchimp says the average campaign bounce rate is between 1% and 2%, that above 2% you should look at your list, and that above 5% you should review it urgently. That guidance is for opted-in marketing email.
• HubSpot can suspend marketing email for an account whose hard bounce rate goes over 5% in a month (and over 0.1% for spam reports).

Cold lists decay as people change jobs, so bounces creep up on any list that isn't re-verified. Verify every list before it goes into a campaign, remove hard bounces straight away, and treat an address that soft-bounces repeatedly as invalid. Yahoo's guidance is the same in plain words: monitor hard and soft bounces and remove invalid recipients promptly.

Authentication: 100%, not "most"

Authentication is pass or fail, so the benchmark is every message. Since February 2024 Gmail requires all senders to set up SPF or DKIM and bulk senders (5,000+ messages a day to personal Gmail accounts) to set up SPF, DKIM and DMARC, with the From domain aligned to SPF or DKIM. Yahoo's requirements match. Since May 5, 2025 Outlook.com has required SPF, DKIM and DMARC (at least p=none, aligned) from domains sending more than 5,000 messages a day to its consumer addresses, and rejects non-compliant mail with 550 5.7.515 (Microsoft's announcement).

Cold email senders rarely hit 5,000 a day on one domain, but meeting the bulk rules anyway costs nothing and removes a reason to filter you. To check, send a message to a Gmail address you own, open "Show original" and confirm SPF, DKIM and DMARC all say PASS.

On Winnr, every domain you buy or connect with nameservers gets SPF, DKIM and DMARC published automatically, with DMARC at p=reject by default. You can confirm the records with the free DMARC checker or DNS lookup.

Inbox placement: measure it, don't borrow it

We are not aware of a public inbox placement benchmark for cold email that we would trust. Placement depends on the provider, the recipient's own history with you, your copy and your list, so a single industry number would not tell you much about your own mail. Measure it instead:

1. Get test addresses at the providers your prospects actually use (Google Workspace and personal Gmail, Microsoft 365 and Outlook.com, Yahoo, and any corporate gateways you see often). A seed testing service gives you a ready-made list.
2. Send a normal message from a real sending mailbox, with the copy you plan to use.
3. Record the folder at each provider: inbox, another tab, or spam.
4. Repeat after big changes to volume, copy, domains or sending tool.

A warming dashboard is not a substitute. Winnr's warming health score, for example, measures how much warmup mail landed in spam inside the warming network over 14 days. It is a useful health signal, but it is not inbox placement at Gmail or Microsoft. How warming works explains the difference.

Reply rate: the outcome metric

Reply rate is the number that tells you whether cold email is working, and unlike opens it isn't distorted by privacy features. The most useful public reference we found is Instantly's Cold Email Benchmark Report 2026, built from data across its platform from January 1 to December 18, 2025. As of October 2026 it reports:

• Average reply rate: 3.43%. Instantly defines reply rate as all replies received (including replies to follow-ups) divided by total emails sent.
• Top quartile: 5.5% or more. Top 10%: 10.7% or more.
• 58% of replies came from the first email in a sequence, and 42% from follow-ups.
• Sequences of 4 to 7 steps performed best, and the best-performing first emails were under 80 words.

Two cautions. These are one platform's users, and every vendor defines reply rate slightly differently (some exclude auto-replies, some divide by contacts rather than emails). Compare your own campaigns against each other first, and use outside numbers only as a rough reference.

Volume per mailbox: 10 to 20 cold emails a day

There is no provider-published "safe" cold volume. Hard limits exist (Google Workspace allows 2,000 messages a day per user, Exchange Online 10,000 recipients a day; see sending limits by provider), but those are ceilings for normal business use, not cold email targets. Cold email infrastructure providers each publish their own guidance:

• Winnr: 10 to 20 cold emails per mailbox per day (15 by default), and 50 or fewer in total once warming is counted.
• Maildoso: up to 15 cold emails per mailbox per day on every plan, per its pricing page as of October 2026.
• Mailforge: recommends 30 emails per mailbox per day (100 maximum) with 2 to 3 mailboxes per domain, per its pricing page as of October 2026.

Scale by adding mailboxes and domains, not by pushing each mailbox harder. The cold email calculator turns a daily target into mailboxes, domains and cost.

Blocklist status: zero listings on the lists that matter

The benchmark here is simple: your sending domains should not be on Spamhaus DBL or SURBL, the domain blocklists most widely checked. Look a domain up on the operator's own site (for example Spamhaus's domain reputation checker) one domain at a time. If a domain is listed, pause it and replace it rather than waiting on a delisting; see how to recover a burned domain.

A weekly checking routine

Winnr handles the infrastructure side of this list: authentication on every domain, built-in warming at $0.60 per mailbox per month, and a 90% deliverability guarantee (if your mailboxes aren't reaching 90% deliverability, ask and Winnr refunds your most recent month). See Winnr pricing for plans.

Sources

All checked October 6 to 7, 2026.

• Google, Email sender guidelines and Email sender guidelines FAQ.
• Yahoo, Sender best practices.
• Microsoft, Outlook's new requirements for high-volume senders.
• Instantly, Cold Email Benchmark Report 2026 (platform data, January 1 to December 18, 2025).
• Mailchimp, How to reduce email bounce rates.
• HubSpot, Why is my email suspended.

Related guides: Cold email best practices, sending limits by provider, which TLDs to use and how many domains you need.