How to benchmark your email program against the industry
Email marketing is at a crossroads. Mailgun's 2026 Email Impact Report, which analyzed more than 400 billion emails sent through its platform in 2025 alongside survey responses from over 1,200 senders, found a widening gap between programs. The programs pulling ahead share three traits: they take deliverability seriously, they measure what actually matters, and they have moved past experimenting with AI.
The gap is less about budget than about practice. Here is how to benchmark your own program against what the data shows.
Why deliverability is your first priority
Start with the report's most uncomfortable finding. 36% of senders say they monitor an "email deliverability rate" — a metric that does not accurately exist. Delivery rate reflects whether a receiving server accepted your message. It says nothing about whether the message reached an inbox. More than one in three teams is monitoring a number they believe means something it does not.
The corollary: only 25% of senders consistently run inbox placement tests, which is the measurement that would tell them the truth. Deliverability is the foundation of any email program, and three quarters of senders are not measuring it.
Industry delivery benchmarks from the report:
Industry | Delivery rate |
|---|---|
Air Freight & Logistics | 99.25% (0.01% bounce rate) |
Media | 95.95% (lowest of the top 10) |
Those four percentage points represent a very different program health picture at scale.
What to do about it. Monitor inbox placement, not delivery rate. Validity's Sender Score gives you a free read on sender reputation, and Validity Everest is the paid platform for ongoing inbox placement testing. If you encounter older guidance recommending Return Path, note that Validity acquired it in 2019 and retired the brand — Return Path Certification is now Validity Sender Certification.
Clean your lists on a schedule rather than reactively, and confirm your SPF, DKIM, and DMARC records are correctly configured. On that last point, the report found DMARC adoption rose from 43% in 2023 to 61% in 2025 — but 27% of DMARC users do not know which policy their own domain publishes. A DMARC record set to p=none monitors and enforces nothing, which is a different situation from having no record at all only in the sense that you have the option to turn it on.
A note on comparing reports: other 2026 deliverability studies have found DMARC adoption declining. The samples differ — Mailgun measures its own sending base, while studies drawn from diagnostic tools skew toward senders already investigating a problem. Read any single figure as a signal about that population rather than about all email.
Measurement: the gap hiding in plain sight
The report's second finding is that most programs cannot prove their own value. 78% of senders rate email as very or extremely important to their business. Fewer than half track its ROI — 46% for promotional email, 43% for transactional.
That matters because the senders who do measure find numbers worth reporting. Among those tracking ROI, 60% report more than $10 returned per $1 spent on promotional email, and 13% report more than $40. The transactional figure is similar at 62%.
Meanwhile, 43% of senders cite budget constraints as the main barrier to investing more in email. Those two findings sit next to each other: the channel most likely to be under-funded is also the one least likely to be documenting its return.
What to do about it. Set up conversion tracking that connects sends to revenue, not just to clicks. Then segment: analyze which audiences respond to which content, and build targeted segments from the behavior rather than from demographics. If you are among the 54% not tracking ROI, that is the single highest-leverage change available to you, and it is a reporting problem rather than a marketing one.
AI: past the experiment, short of the default
79% of senders use or plan to use AI in email. Only 27% use it regularly. The gap between intention and habit is where most programs currently sit.
For those who have made it a habit, the results support the investment: 54% of AI users reported program improvement in 2025, against 37% of non-users. That is a real seventeen-point spread.
But the honest version includes the other number. 23% of AI users reported no benefit at all. AI is not a magic bullet, and nearly a quarter of adopters have the data to prove it.
The difference between the 54% and the 23% is usually application rather than tooling. Subject line optimization — tools like Scalero's AI subject line generator — works because the task is pattern recognition against a measurable outcome. Predictive send timing works for the same reason. Applications where AI replaces judgment rather than accelerating analysis are where the 23% tends to cluster.
Benchmarking your program: a practical approach
Assess your current state in each of the three areas, using the report's figures as the comparison:
Deliverability. Are you measuring inbox placement or delivery rate? If the latter, you are in the 36%. Do you run placement tests consistently? If not, you are in the 75%. Check your DMARC policy value, not just its existence.
Measurement. Can you state your email ROI with a number? If not, you are in the 54% who cannot, and you are likely also in the 43% citing budget constraints — those are related.
AI. Do you use AI regularly, or do you intend to? If the latter, you are in the 52-point gap between the 79% who plan and the 27% who do. Start with one application that has a measurable outcome.
The pattern across all three is the same: the gap is rarely between teams that know what to do and teams that do not. It is between teams that measure and teams that assume.
The bottom line
Mailgun's data describes a widening divide, and the dividing line is measurement. Programs that know their inbox placement, their ROI, and their AI results can improve deliberately. Programs working from delivery rates, open rates, and intentions are optimizing against numbers that do not describe their situation.
Pick the one area where you cannot currently state a number, and fix the measurement before you change the tactics.
For the full dataset, see the Mailgun 2026 Email Impact Report coverage on Marketer's Index. If you want a second set of eyes on where your program sits, our lifecycle marketing services include program audits across all three areas.
