Top ESPs for French DTC Brands: 2026 Overview

Every quarter, the same question comes up in audits: "what's the best ESP in France today?" It's the wrong question. There is no official ranking of ESPs on the French market, no market share published by vendors, no comparative deliverability figure measured on a common basis. Whatever circulates as a podium is either opinion or sponsored content.

What is measurable, on the other hand, is the billing model, the documented technical constraints, and compatibility with your stack. These three angles are enough to settle 90% of cases. This overview goes through them, with the official source behind every figure.

What a "top ESP" list can't tell you

Before getting into the detail, let's set the limits of the exercise.

No vendor publishes a deliverability measure comparable to a competitor's. The inbox rates shown in sales decks come from unaudited internal methodologies: they don't compare to one another. Deliverability depends on your authentication, your list hygiene and your sending cadence far more than on the logo on the invoice. Our anti-spam deliverability guide details this split of responsibilities.

The number of French brands on any given platform isn't published either. Neither is market share. An article that claims "X is number one in France" is either inventing the figure or recycling an unsourced study.

What remains verifiable: public pricing grids, billing units, API limits, and inbox provider requirements. That's where the decision actually gets made.

The first criterion: what the ESP actually bills

Two platforms with a similar sticker price can produce very different invoices, because they don't count the same thing. Three families of billing models coexist.

Billing by reachable profile

Klaviyo bills the active profile. The official definition is broad: any profile that can be emailed in Klaviyo counts as an active email profile, regardless of consent status (Klaviyo documentation). The billable base covers profiles that "can receive at least one type of marketing, have a valid email address or phone number, and are not suppressed or deleted" (source).

Direct operational consequence: cleaning your list lowers the bill. Deleting a profile or suppressing it for good takes it out of the billable base. That's why a properly wired sunset flow pays for itself twice over: on sender reputation and on the spend line. The method is detailed in our Klaviyo list hygiene article.

Klaviyo's free plan is capped at 250 profiles and 500 emails a month (Klaviyo pricing page). Fine for a technical test, not for a commercial launch.

Billing by send

Shopify Messaging works the other way around: the meter is the message sent, not the contact stored. The published grid is clear (Shopify help):

Two details matter a lot in this model. Abandoned cart automations don't draw on the free monthly quota, unlike other automations. And unused credit is lost at the end of the calendar month: there's no way to bank volume in September to absorb the November spike. A highly seasonal brand therefore pays full price during its Black Friday weeks, with no offset during its slow months. Our Black Friday email checklist builds this constraint into the send plan.

Billing by identified profile or active user

Customer.io bills the unique identified profile, people and objects combined, with 5,000 profiles included on the Essentials plan and $0.009 per additional profile (Customer.io pricing page). The notion of "object" matters: accounts, workspaces, subscriptions, anything you model counts toward the base. A SaaS product that creates one object per organization inflates its bill without adding a single human recipient. The higher tiers (Standard, Premium, Enterprise) run on custom volumes and pricing that aren't published.

Braze bills by monthly active user (MAU), described as "the customers who actively interact with your brand across your digital experiences," and publishes no price at all (Braze pricing page). It's a quote-based model, positioned above the standard ecommerce ESPs. Any price range you find online for Braze is a reconstruction, not vendor data. Our Braze vs Klaviyo comparison digs into the cases where that level of investment is justified.

On Brevo's side, pricing and free-plan limits circulate widely, but the official pages don't render in a form our verification tools can read. So we're not repeating any figure here: the free plan's daily email count, the contact cap, and the exact names of the 2026 paid plans depend solely on the vendor's own pricing page, the only reliable source on this point. Our detailed pricing-by-contact-volume article lays out the calculation method, independent of the displayed amounts.

What this changes in your modeling

Before signing, calculate your cost over a rolling twelve months, not an average month:

• Billing by contact: your cost tracks list growth, even if you send nothing in July.
• Billing by send: your cost tracks your promotional calendar, with sharp spikes in Q4.
• Billing by MAU or identified profile: your cost tracks product activity, not marketing activity.

A brand that captures aggressively via pop-up but sends little will pay more under a per-contact model. A brand with a stable list that sends five times a week will pay more under a per-send model. Calculating subscriber value lets you compare that spend against the revenue actually generated.

The second criterion: technical integration constraints

Price gets negotiated once. Technical limits, you live with every day.

API limits

Brevo's API is capped in requests per second and per hour, with the ceiling depending on your plan: the advanced limits are reserved for Professional and Enterprise accounts (Brevo developer documentation). Exceeding the quota returns an HTTP 429 Too Many Requests code, which forces a retry logic on the integration side.

This matters a lot if you push events from a custom back office, a PIM, or a data warehouse. An initial import of several hundred thousand contacts, a real-time order sync, or a poorly spaced nightly batch will all run into this ceiling. Plan for exponential backoff at design time, not after your first incident. The equivalent logic on Klaviyo's side is described in our Klaviyo Events API guide.

Depth of the ecommerce integration

A native integration isn't judged by the presence of a connector, but by what it actually pulls in:

• Product events (viewed, added to cart, checkout started), not just orders.
• A synced catalog, a prerequisite for dynamic product blocks.
• Order statuses, a prerequisite for post-purchase flows conditioned on delivery.
• Historical customer properties, a prerequisite for RFM segmentation.

Without a synced catalog, there's no serious browse abandonment. Without delivery status, there's no post-purchase sequence timed to actual receipt. That's the difference between a send tool and a lifecycle platform, and it's what decides your ceiling on owned revenue. Our Klaviyo Shopify integration guide lists the objects to check at connection time.

What this overview can't verify

Two topics come up systematically in RFPs and deserve an honest answer: data location (EU residency, standard contractual clauses) and the compliance certifications claimed (ISO 27001, SOC 2, DPA). These vary by vendor, by plan and by date. None of them appear on any of the official pages we were able to open for this overview. Demand the vendor's contractual documentation before signing, especially if you handle health data or data on minors.

The third criterion: your ability to meet inbox provider requirements

An ESP doesn't get you to the inbox. It gives you the tools to meet rules that are set by Gmail and Yahoo. Those rules apply no matter which tool you use.

Google imposes stricter requirements on senders of more than 5,000 messages a day to Gmail accounts (Google requirements). That threshold is crossed very quickly: a list of 60,000 contacts sent to weekly clears it on every single campaign.

The points to validate, in order:

• DMARC alignment: the From field's domain must align with either the SPF domain or the DKIM domain (Google).
• SPF and DKIM combined on Yahoo's side, plus a valid DMARC policy at minimum p=none (Yahoo best practices).
• Complaint rate under 0.10% in Postmaster Tools, never reaching 0.30% (Google). Yahoo asks senders to stay under 0.3% (Yahoo).
• One-click unsubscribe for marketing messages, in addition to a visible unsubscribe link (Google).
• Unsubscribe processing within 2 days (Yahoo).

DMARC rollout happens in stages: publish a record with the none policy first to collect reports, then move to quarantine, then to reject (dmarc.org). Jumping straight to p=reject without reading the reports means blocking your own transactional flows. The full DNS procedure is in our SPF, DKIM, DMARC article, and the 2026 state of play in Google and Yahoo sender requirements.

These requirements don't depend on the ESP. What does depend on it is how easy they are to configure: dedicated sending domain management, a compliant unsubscribe header handled natively, access to bounce logs, granularity of suppressions. Ask these four questions in the demo.

The fourth criterion: the French legal framework

This topic often gets treated last, even though it governs how the list itself gets built.

For companies operating in France, prospecting individuals by email requires consent collected before the solicitation, evidenced by a positive action such as an unchecked checkbox (CNIL). The exception for an existing customer is narrow: it only covers similar products or services supplied by the same company (CNIL). Selling sneakers doesn't authorize you to prospect the same list for a range of supplements under a different brand in the group.

Every solicitation must let the person express, by a simple means, their refusal to receive further solicitations (CNIL).

What this means for tool choice: the ability to timestamp and store proof of consent, to distinguish email consent from SMS consent, to manage several lists with distinct legal bases. A platform that overwrites the consent history on every import puts you at risk. Our GDPR compliance checklist goes through the control points one by one.

The decision grid, in six questions

Rather than a ranking, here's the sequence we apply in audits. Each answer eliminates candidates.

1. What's your dominant cost unit? A list that grows fast, or a send volume that explodes in Q4. The answer points you toward a per-contact or per-send model.
2. How deep does your data need to go? Orders alone, or product events and custom objects. A need for custom objects disqualifies purely ecommerce tools.
3. Who's wiring the integration? Without a technical resource, an API with variable quotas and manual retries is a wall.
4. What's your daily volume to Gmail? Past 5,000 messages a day, the stricter requirements apply, and the tool needs to cover them natively.
5. How many legal bases coexist in your list? Multi-brand, multi-country, and mixed B2B force fine-grained consent management.
6. What's the exit cost? Exporting profiles, historical events, and templates. A poorly prepared migration costs more than a year of subscription, as our Brevo to Klaviyo migration guide shows.

A Shopify DTC brand with 40,000 contacts and a rich catalog, a multi-product B2B SME, and a product-led SaaS won't converge on the same answer. That's normal: they don't share the same dominant cost unit or the same data depth.

What matters more than which ESP you pick

An agency observation, repeated across dozens of accounts: the performance gap between two brands on the same tool is almost always bigger than the gap between two tools for the same brand.

The levers that actually carry weight:

• The number of flows genuinely active, and their coverage of the lifecycle.
• The quality of the segmentation, not the number of segments created.
• List hygiene discipline, which protects your reputation and lightens the per-contact bill.
• A send cadence held over twelve months, not over six weeks in Q4.

Our 50-point lifecycle scorecard measures these four axes before considering any platform change. In most audits, the existing account still has significant room to improve before the tool becomes the limiting factor.

If you've already settled on the platform and the topic is execution, our Klaviyo agency and our Brevo agency work on both stacks daily.

FAQ

Is there an official ranking of ESPs in France?

No. No vendor or regulatory source publishes market share, French customer counts, or a ranking of ESPs on the French market. The podiums you come across are editorial opinions or sponsored content.

Can deliverability rates be compared across platforms?

Not reliably. No vendor publishes a methodologically comparable measure. The indicators worth watching are your own: complaint rate under 0.10% in Postmaster Tools and never above 0.30% per Google, under 0.3% per Yahoo.

Does cleaning your list really lower the Klaviyo bill?

Yes, when billing is based on active profiles. Klaviyo counts profiles that can receive at least one type of marketing and are not suppressed or deleted (documentation). Taking a profile out of the billable base mechanically reduces the billed volume.

Is the per-send model cheaper than the per-contact model?

It depends on your sending profile. Shopify offers 10,000 free emails a month then charges $1 per 1,000 up to 300,000 (Shopify help), with credit that doesn't carry over month to month. A highly seasonal brand therefore absorbs its spikes poorly. Model it over twelve months before deciding.

Do you need a paid plan to integrate Brevo with a custom stack?

API limits depend on the plan: the advanced limits are reserved for Professional and Enterprise accounts, and exceeding them returns an HTTP 429 code (Brevo developer documentation). If your integration pushes large volumes of events, check your plan's ceiling before you build.

Weighing an ESP switch or a first setup for a French DTC brand? Talk to the Deliver team.