The line about email returning thirty-six dollars for every dollar spent has been repeated so often that almost nobody checks where it came from any more. It came from a marketer survey, which is a very different object from a measured result. That distinction matters, because the case for email in 2026 is stronger than the slogan and rests on something the slogan never mentions.
Key takeaways
- The DMA’s Marketer Email Tracker, published 25 March 2026, put ROI at roughly £41 per £1 spent, up from about £38.
- Litmus’s 2025 State of Email survey of nearly 500 marketers reported sector returns from 32:1 to 45:1.
- Gmail and Yahoo have enforced authentication and one-click unsubscribe for bulk senders since February 2024.
- Apple’s Mail Privacy Protection has made open rate unusable as a performance metric.
Where the ROI figures come from, and what they are worth
Two trackers dominate the citation trail, and both are surveys of marketers rather than audits of accounts. The DMA in the UK published its Marketer Email Tracker 2026 on 25 March 2026, reporting that return per pound spent had risen from around £38 to about £41. Litmus, in its 2025 State of Email survey of nearly 500 marketing professionals worldwide, broke the number down by sector: media, publishing and entertainment at 32:1, software and technology at 36:1, agencies at 42:1, and retail, ecommerce and consumer goods at 45:1.
Read those as directional. They tell you the ranking of email against other channels holds up across markets and sectors, which is genuinely useful. They do not tell you what your programme will return, and the spread between 32:1 and 45:1 in the same survey is the honest warning label. Litmus also reported that 21% of marketing leaders still do not measure email ROI at all, down from 36% in 2023, which means a meaningful slice of the industry is quoting a number it has never calculated for itself.
The site’s own position on this is unchanged: we do not treat a survey average as a forecast for anybody’s programme, and neither should a board deck.
The advantage nobody puts in the slide: an owned list behind an authenticated gate
What actually separates email from paid social is not conversion rate. It is that nothing sits between the sender and the recipient that can reprice or deprioritise the relationship overnight. A subscriber list survives a platform’s algorithm change, an ad account suspension and a CPM spike.
Since 2024 that gate has hardened in a way that quietly favours anyone doing it properly. The major mailbox providers now impose authentication and unsubscribe requirements on high-volume senders, and non-compliant mail no longer merely underperforms.
| Provider | In force since | What bulk senders must have |
|---|---|---|
| Gmail | February 2024 | SPF, DKIM and DMARC; one-click unsubscribe (RFC 8058); spam complaint rate under 0.3% |
| Yahoo | February 2024 | Aligned with Gmail on authentication and one-click unsubscribe |
| Microsoft (Outlook, Hotmail, MSN) | 5 May 2025 | SPF, DKIM and DMARC at minimum p=none; failing mail rejected outright rather than junked |
The threshold in every case is 5,000 messages a day to that provider’s consumer domains. A brand sending to a list it bought will trip the complaint ceiling before it trips anything else. A brand sending to people who asked will not notice the rules exist.
Open rate stopped being a metric in 2021
Any measurement conversation has to start by retiring the number most reports still lead with. Apple’s Mail Privacy Protection, introduced with iOS 15 in 2021, preloads remote content whether or not the recipient opens anything, which means the tracking pixel fires regardless. Litmus put Apple at roughly 49% of all email opens in January 2025, so this is not an edge case affecting a slice of the file.
The practical consequence is that open rate now measures a mix of real opens and machine prefetches in unknown proportions, and it inflates. Clicks, replies, and revenue attributed to a send remain measurable because MPP does not touch them. Any subject-line test judged on opens after 2021 was measuring noise.
Consent is what decides whether the programme is legal
The commercial argument collapses if the legal basis is wrong, and the rules differ by market rather than by mailbox. Anyone sending across borders needs to know which regime applies to each portion of the list.
Under the GDPR and the ePrivacy rules that implement it nationally, marketing email to individuals in the EU and the UK generally requires prior opt-in consent, freely given and recorded, with a narrow soft opt-in for existing customers being sold similar products. Consent has to be as easy to withdraw as it was to give, and the record of when and how it was obtained is the thing regulators ask for first.
Under the US CAN-SPAM Act the logic is inverted: prior consent is not required, but every commercial message must carry a valid physical postal address, must not use deceptive headers or subject lines, and must honour an opt-out request within ten business days. One-click unsubscribe, now a mailbox provider requirement as well, sits comfortably inside both regimes.
Where a list mixes jurisdictions, the workable answer is to apply the stricter standard across the file rather than to segment consent logic by country and hope the mapping holds. These frameworks also move: verify the current obligations in each market you send to before a campaign ships, not after.
Four objections we hear, and what the evidence says
- “Inboxes are saturated.” Complaint rate is the metric that matters, not volume, and it is entirely under the sender’s control. A file that opted in and gets relevant mail sits far below the 0.3% ceiling.
- “Younger audiences don’t use email.” They use it as an account and transaction layer rather than a conversation layer, which changes format and cadence rather than viability. That is a design problem, and it is the same one we unpacked in why your funnel stopped working.
- “The ROI number is too good to be true.” Partly, yes. It is self-reported and unaudited. The relative ranking against paid channels is the defensible part.
- “Deliverability has become too technical.” It has become more demanding and more standardised at the same time. SPF, DKIM and DMARC are a one-off configuration, not an ongoing tax.
What we would put on the dashboard instead
- Revenue per subscriber per month, which forces list quality and send frequency into the same number.
- Click-to-conversion rate, unaffected by Mail Privacy Protection and closer to the outcome anyone actually cares about.
- Complaint rate against the 0.3% ceiling, tracked per campaign type rather than as a monthly average.
- Authenticated delivery share, verified through DMARC reporting rather than assumed from the ESP dashboard.
- List decay rate, because a programme that returns 40:1 on a shrinking file is borrowing from next year.
Email is underrated less because people doubt it works and more because its advantages are unglamorous. Ownership, authentication and permission are not campaign ideas. They are the reasons a channel is still standing after twenty years of being declared finished, and they are the only part of the story worth putting in front of a finance director.
Know what to fix before you scale it?
Most underperforming programmes fail on a short list of avoidable errors rather than on strategy.
Sources: DMA (UK) Marketer Email Tracker 2026, published 25 March 2026; Litmus State of Email survey 2025, nearly 500 marketing professionals; Litmus Email Client Market Share, January 2025; Google Workspace and Yahoo bulk sender guidelines, in force February 2024, including RFC 8058 one-click unsubscribe; Microsoft Outlook high-volume sender requirements, in force 5 May 2025; GDPR and ePrivacy rules, and the US CAN-SPAM Act. ROI figures are self-reported marketer survey data and should be read as directional, not as a forecast. Legal obligations vary by market and change: confirm current requirements before sending. Updated August 2026.
