Newsletters often get lumped in with spam: that promotional email nobody asked for and deletes without reading. But the technical definition is much stricter than that. A newsletter is an informational email sent at regular intervals to a list of contacts who have given explicit consent (opt-in). It differs from a one-off email marketing campaign through its announced cadence and editorial content. In 2026, the format remains the channel with the best measured return on investment: between $36 and $42 generated per dollar spent, according to the State of Email 2025 report by Litmus and the DMA. The question is what the numbers displayed by an email tool actually tell you.
The criteria that define a newsletter
Four elements are enough to identify a newsletter: an opt-in contact list, a frequency announced in advance, anywhere from daily to monthly, editorial content rather than pure promotion, and an always-accessible unsubscribe link. A transactional email (order confirmation, password reset) meets none of these criteria and doesn’t need marketing consent to be sent. A newsletter belongs to the realm of inbound marketing: it informs an audience that chose to receive it, rather than interrupting an unsolicited prospect.
Newsletter vs. email marketing: the distinction in brief
A newsletter is a subcategory of email marketing: it inherits the same sending tools and the same consent requirements, but it follows an announced cadence and plays an editorial role. Email marketing, in the broader sense, also covers one-off promotional campaigns, abandoned cart follow-ups, welcome sequences, or restock alerts: sends triggered by behavior, with no fixed schedule. This distinction shapes content strategy as much as performance measurement; it’s covered in detail in our complete comparison of newsletters and email marketing.
What the performance metrics actually measure
A 45% open rate looks impressive, until you look at who measured it. The average across France hovers around 18.2% across all sectors, a figure published by DMA France and cited in the Klaviyo 2025 benchmarks; it climbs to 26% when the subject line is personalized. This rate stays a noisy indicator: some of these opens come from mail servers themselves, without any human clicking anything. The click-through rate holds up better against this pollution, since it only triggers on an actual action by the recipient. An open rate that swings sharply without any content change to explain it is usually the symptom of a measurement problem rather than an audience one.
| Metric | Measured value (2025) | What it hides |
|---|---|---|
| Average open rate (France) | 18.2% across all sectors (DMA France / Klaviyo, 2025) | Includes automatic opens triggered by Apple Mail |
| Open rate with personalized subject line | 26% (DMA France, 2025) | Real personalization effect, on the same measurement base |
| Average email marketing ROI | $36 to $42 generated per dollar spent (Litmus / DMA) | Measured over the full cycle, beyond a single send |
| Best-performing send window | Tuesday, between 10am and 11am (Mailjet, 2025) | Average across all sectors, to be recalibrated for your own audience |
Opt-in rules before the first send
A list rented or bought with one click, imported without any verification: the entire newsletter inherits the problem from the very first send. The framework to follow before that first send comes down to four points:
- Collect explicit, tracked, and dated consent: a pre-checked box doesn’t count as valid consent under GDPR, a position the CNIL has held consistently since 2018.
- State the sending frequency of the newsletter at signup, to avoid unsubscribes driven by surprise rather than content (see our guide on sending frequency).
- Keep proof of consent (timestamped signup form and the contact’s IP address) in case of an audit.
- Display a visible unsubscribe link in every send, with no extra confirmation step.
The full legal framework, including CNIL penalties, is covered in our dedicated article on GDPR applied to email marketing. A list built outside these rules can’t be fixed after the fact: purchased or rented addresses generate a complaint rate that damages sender reputation from the very first sends. A list that’s compliant on paper doesn’t yet guarantee a real audience: the metrics that follow the send are what settle that question.
Why the open rate can lie
An e-commerce newsletter jumps overnight from a 28% to a 55% open rate, with no change to the subject line or send time. This scenario has a name: Apple Mail Privacy Protection, documented by Paubox in 2025. The feature, active on Mail since 2021, automatically preloads the content and tracking pixels of every message received, before the recipient even opens the app. The sender’s server then logs an open, whether or not a human actually read the message.
According to an analysis published by Paubox in 2025, this mechanism overstates measured open rates by 15 to 20 points compared to recipients’ actual engagement.
Apple Mail accounts for a large share of email opens worldwide, which makes the effect impossible to ignore on a general consumer list. This bias is still the blind spot of most email dashboards. The click-through rate largely escapes this technical bias. That’s why more and more marketing teams track CTOR (click-to-open rate, the ratio of clicks among opens) rather than the raw open rate to judge a piece of content’s real relevance. That national average of 18.2% hides a well known sector gap: low-volume B2B newsletters often post noticeably higher rates, while large-list e-commerce newsletters post noticeably lower ones. The average erases both.
Automation, segmentation, and frequency: what keeps a newsletter going over time
A newsletter that lasts more than a year rarely relies on a single manual send. Most teams automate at least the welcome message and re-engagement of inactive contacts, through automated sequences (flows at Klaviyo, workflows at other tools), triggered by signup or by a lack of opens over a given period. The segmentation feeding these sequences often relies on RFM analysis (recency, frequency, monetary). It mainly separates contacts who are still active from dormant ones who haven’t clicked in months, two profiles that shouldn’t receive the same sending rhythm.
The contact base itself degrades mechanically over time: closed mailboxes, mistyped addresses, duplicates never cleaned up, catchall accounts that accept everything without ever reading it. Every unhandled hard bounce weighs on sender reputation, tracked by Gmail Postmaster Tools and its equivalents at Microsoft and Yahoo. Feedback loops set up by mailbox providers also report spam complaints directly to the router, before the recipient even unsubscribes. Past a certain rejection threshold, the sending IP address gets flagged for monitoring, and so does the associated domain: even messages meant for active contacts start landing in spam. A list never verified before sending accumulates these dead addresses with no warning signal before deliverability collapses. Filtering out bounces after the send fixes nothing, since the message has already gone out and the reputation has already shifted: the only moment verification changes anything is before you hit send.
Subject line and mobile rendering: the details that decide a click or a scroll
Mobile rendering is no longer a secondary variable: the majority of newsletter opens now happen on smartphones, and a non-responsive email (columns that overflow, text unreadable without zooming) gets closed in under three seconds. The subject line remains the first filter: it decides whether the message gets past the preview or ends up ignored in the list. The choice of topic often follows the recommendations set by the campaign manager, built around a precise editorial calendar rather than improvisation. Testing two versions of the same subject line on a small sample before the full send, A/B testing applied to the subject line, stays the most reliable method for choosing between two options, better than an unverified internal preference.
These tactical settings vary from one list to the next. The format itself hasn’t changed in nature since its early days on the web: informing an audience that said yes, at a pace it already knows. The measurement tools, on the other hand, are still learning to count accurately.
