The Cold Email Deliverability Benchmark
Updated June 17, 2026
Cold-email results are gated by deliverability before copy ever matters. Modeling inbox placement across configurations, a fresh unauthenticated domain lands roughly 25% in the inbox while a warmed domain with SPF, DKIM, and DMARC lands near 90%. Because replies are a product of inbox placement, the well-configured sender gets over 3x the replies from the identical message and list.
Operators blame their copy when cold email underperforms, but the message is usually irrelevant — it never reached the inbox. Deliverability is decided upstream by domain age, warmup, and authentication, and an email in the spam folder converts at zero no matter how good the subject line is.
This is a transparent model, not a survey. Reply yield is computed as emails × inbox-placement rate × in-inbox reply rate, with placement rates assigned to each configuration tier and every assumption stated so you can reproduce it in the deliverability calculator.
Methodology
Reply yield = emails sent × inbox-placement rate × in-inbox reply rate. The table fixes volume at 1,000 emails and the in-inbox reply rate at 3% (replies among messages that actually land in the inbox), then varies the placement rate by configuration tier.
Placement rates are modeled per tier: fresh + no auth at 25%, fresh + full auth at 55%, warmed + partial auth at 75%, and warmed + full auth (SPF, DKIM, DMARC) at 90%. These reflect the documented gap between unconfigured and fully configured senders, not measured values for any specific provider.
The 3% in-inbox reply rate is held constant across tiers on purpose — it isolates deliverability as the variable. Holding copy and list quality fixed, the model shows what configuration alone does to outcomes.
The data
| Configuration | Inbox placement | Emails in inbox | Modeled replies | Replies vs. baseline |
|---|---|---|---|---|
| Fresh domain, no auth | 25% | 250 | 7.5 | 1.0x |
| Fresh domain, full auth | 55% | 550 | 16.5 | 2.2x |
| Warmed, partial auth | 75% | 750 | 22.5 | 3.0x |
| Warmed, full auth (SPF+DKIM+DMARC) | 90% | 900 | 27.0 | 3.6x |
Replies = 1,000 emails × placement rate × 3% in-inbox reply rate. Placement assigned per configuration tier; reply rate held constant.
Placement, not copy, is the first gate
Because reply yield is a product of placement and reply rate, placement caps everything downstream. A fresh, unauthenticated domain lands about 25% in the inbox, so 75% of the send is invisible before the subject line is even read — and the best copy in the world cannot recover a message in spam.
This is why two senders with identical lists and messaging post wildly different results. The variable that separates them is rarely the words; it is whether the receiving server trusted the sender enough to deliver. Configuration is the lever, and it is fully within the sender's control.
Authentication and warmup stack
The tiers compound. Adding full authentication to a fresh domain more than doubles placement (25% to 55%) by proving the sender is who it claims to be. Warming the domain on top of that pushes placement to 75–90%, because a gradual ramp signals legitimate sending behavior rather than a spam burst.
Neither lever substitutes for the other. Authentication without warmup still looks like a cold cannon; warmup without authentication still fails identity checks. The model shows the top tier — warmed plus SPF, DKIM, and DMARC — landing near 90%, roughly 3.6x the placement of the unconfigured baseline.
The reply gap is the whole game
Translated to replies, the gap is decisive: at 1,000 emails and a 3% in-inbox reply rate, the unconfigured sender earns about 7.5 replies while the fully configured sender earns 27 — from the same message and the same list. That is over 3x the pipeline for zero additional copy or list spend.
The takeaway is that deliverability is the highest-leverage fix in cold email, and it is purely technical. BILT handles domain setup, authentication, and warmup so the configuration tier is the top one by default — which is why the same campaign produces multiples more replies through it than through an unconfigured stack.
Frequently asked
What matters more in cold email — copy or deliverability?
Deliverability comes first. The model holds copy constant and shows replies tripling purely from configuration: a fresh unauthenticated domain lands 25% in the inbox versus 90% for a warmed, fully authenticated one. A message in spam converts at zero regardless of how good it is.
How much does authentication actually help?
In the model, adding SPF, DKIM, and DMARC to a fresh domain lifts inbox placement from 25% to 55% — more than double — because it proves sender identity to the receiving server. Warmup stacks on top, pushing placement to 75–90%. Neither lever substitutes for the other.
What's a realistic reply rate from a well-configured sender?
The model uses a 3% in-inbox reply rate held constant across tiers to isolate deliverability. At 1,000 emails, that's about 27 replies at 90% placement versus 7.5 at 25% — same copy, same list, over 3x the result from configuration alone.
Are these placement rates measured?
No — they are modeled per configuration tier (25/55/75/90%) to reflect the documented gap between unconfigured and fully configured senders, not values for any specific inbox provider. Reproduce them with your own placement assumptions in the deliverability calculator.
The takeaway
Cold email is won or lost on deliverability before copy matters. The model holds the message and list fixed and shows replies climbing from 7.5 to 27 per 1,000 emails purely by moving from a fresh, unauthenticated domain to a warmed one with SPF, DKIM, and DMARC. Configuration is the highest-leverage and most controllable lever in the channel — fix the inbox-placement gate first, then worry about the words.