Deliverability decoded: Pierre Pignault on what really impacts your inbox placement in 2025

Email deliverability is no longer just about good content or catchy subject lines — it’s about precision, strategy, and staying ahead of ever-evolving spam filters. That’s why we spoke with Pierre Pignault, a leading expert in the field and founder of MailSoar, to unpack what really makes or breaks email performance in 2025. From subdomain strategies and smart segmentation to the realities of AI-generated content and Gmail tab placement, Pierre shared a practical and eye-opening perspective that every email marketer should hear.

Pierre Pignault is the founder of MailSoar, an email deliverability consulting agency that works with senders around the world to secure and optimize their email performance. MailSoar specializes in performing infrastructure audits, implementing deliverability best practices, and offering long-term support for companies looking to build email systems that last. Whether you’re sending marketing emails, transactional messages, or cold outreach, Pierre and his team ensure that your emails land where they should — safely in the inbox.

From metrics to messaging: Deliverability tactics that actually work in 2025

Stripo: Many email marketers focus on open rates, but inboxing performance is often more nuanced. What are the best ways to track and interpret inbox placement trends, and what KPIs truly indicate success?

Pierre: We typically correlate three metrics to determine inbox or spam:

• unique open rates (the sheer value but also the trend);
• reputation metrics (Google Postmaster, SNDS, and other tools);
• inbox placement test.

Ideally, to make a determination, we need the three insights to go the same way. It can happen that only 2 out of 3 will show consistent metrics. In this case, the experience and talent of the consultant come in handy when assessing whether emails go to spam or the main inbox. Inbox placement tests, for example, can be volatile and show spam, while the other metrics show the main inbox. We can also add more testing by, for example, signing up our own email addresses to the marketing program and seeing where we get emails.

S: Engagement rate is increasingly seen as a key factor in inbox placement. What strategies do you recommend for improving engagement while keeping emails deliverable at scale?

P: We typically set up waterfall segmentation strategies with our clients.

We recommend segmenting your subscribers into three groups:

1. Engaged subscribers (opened an email in the last 3 months) with new signups.

Strategy: This highly engaged group should receive the majority of your email campaigns. You can send emails frequently to support your marketing objectives. Ideally, 80% of your total email volume should be directed at this segment.

Recommended KPIs:

• open rate > 45%;
• bounce rate < 0.2%;
• unsubscribe rate < 0.5%.

2. Disengaged subscribers (opened an email in the last 3–6 months).

Strategy: These recipients show some engagement but risk becoming inactive. Their open rates will naturally be lower, and bounce rates may be slightly higher. To maintain deliverability, limit sending to once a week, spreading volume over a period (e.g., 7 days). This group should represent no more than 15% of your total email volume.

Recommended KPIs:

• open rate: > 15%;
• bounce rate: < 0.5%;
• inactive subscribers (no opens in the last 6+ months).

Strategy: This segment poses the highest risk of spam complaints and bounces. Limit sending to 5% or less of your total volume and space out emails over several weeks or months. If engagement remains low, consider sunsetting these contacts to protect your sender’s reputation.

Recommendation:

• Send emails once a month at most, with gradual volume throttling.

S: In your experience, what’s the biggest misconception about A/B testing in email marketing? How can brands go beyond surface-level optimizations to uncover real improvements in engagement and deliverability?

P: I am more on the deliverability side of things, but I’ll try to answer based on what I’ve seen. Sometimes, A/B testing is meaningless because the marketer does not define the hypothesis that they want to prove or disprove. I would also see some failed A/B tests due to the lack of proper test conditions.

For example, I had a client was A/B testing the length of their emails. However, the test failed to uncover exactly what the ideal length of the email was because introducing longer emails also added more topics. So, in the end, the client didn’t really know whether their subscribers liked longer emails, more topics, or more specialized topics (which is what the client added). I believe that they had to redo several rounds of tests. In the end, shorter, straight-to-the-point, and single-topic-only emails worked better.

S: Cold emailing has become increasingly complex, with stricter spam filters and new regulations. What are the essential elements of a cold email infrastructure that ensure both high deliverability and engagement?

P: Cold emailing has indeed become more difficult, but primarily due to the amount of spam sent under a very vague “legitimate interest” banner. I think we are all annoyed by the number of commercial unsolicited emails that we receive. I am not against cold emailing per se — meaning that I see some innovative companies using this means to reach out to really well-targeted prospects, with a nice message and respecting their inbox. But I also see so many companies targeting really broadly under the “they could be interested” philosophy, which, in my opinion, ruins cold emailing for everyone.

In the last few years, we have also seen the rise of inbox warmers. These tools generate artificial positive engagement on your mail streams to balance negative engagement and maintain deliverability. I’m not exactly sure how these tools will evolve, and spam filters are actively trying to work against these tools and senders. But for now, we have seen them work, so using tools such as these would be a must in one’s 2025 cold emailing toolbox.

Think before you send: Common pitfalls and how to future-proof your email strategy

S: Whether it’s a transactional, marketing, or cold email, setting up a robust email infrastructure requires precision. What are the most overlooked aspects when building a system from scratch, and how do you ensure long-term deliverability and stability?

P: I often see companies using their main domain everywhere and lacking subdomain segmentation. Subdomains are key to any successful deliverability strategy. They allow you to do two things:

1. First, you can protect the reputation of your main domain and other subdomains. Subdomains will have their own “deliverability bubbles.” 
2. It helps for investigation and monitoring purposes. You can set up Google Postmaster on subdomains to track reputation granularly and immediately tie any deliverability issue to its corresponding infrastructure.

Since the Google and Yahoo updates on sender requirements in 2024, I have seen fewer senders with faulty authentication (although it can happen).

You might also like

Adapting to Gmail and Yahoo’s sender requirements: What you need to know

S: What are the most common but underestimated mistakes in email marketing strategy that silently sabotage deliverability? Are there any “innocent” best practices that hurt inboxing?

P: Maybe I would opt to resend emails to non-openers. I have seen a few senders do that lately and end up damaging their deliverability. In one case, their ESP even recommended it, which is a shame. 

Rather, I suggest setting up re-engagement campaigns through automations, with specific conditions to enter the automation (for example, the last open date being beyond the last 180 days). Doing this will ensure that you give your subscribers a chance to re-engage with your content while still maintaining safe segmentation and sending practices.

S: With ongoing changes such as Apple’s MPP, BIMI adoption, and new spam filtering AI, what shifts do you predict in email deliverability over the next two years? How should email marketers prepare for these changes?

P: I think that spam filters will align pretty much with what Gmail does and will require strict authentication protocols to deliver emails. They will also have the ability to analyze the recipient’s behavior and make decisions based on it rather than on more stable variables, such as IP and domain reputation. Finally, I would say that all spam filters are investing in domain reputation-based filtering more than IP. With the rise of shared IP and the possibility of easily switching from one IP to another, the domain remains a significant variable that the spam filter wants to aggregate reputation on.

Balancing performance and placement: How to tackle Gmail tabs, AI personalization, and automated journeys

S: Many marketers struggle with email landing in Gmail’s Promotions or Social tabs. Beyond sender reputation and authentication, what are the most effective content and technical strategies for improving inbox placement?

Something that can be overlooked is that the Promotion tab can also positively impact your deliverability by lowering the spam report and unsubscribe rates. So, in some cases, landing there can help with the overall delivery of the infrastructure. Besides the two points mentioned above, Promotion tab filtering is heavily content-based. 

The approach that we typically use is as follows:

• taking the faulty template and breaking it into smaller pieces (sometimes defaulting to a text-only version); 
• controlling which portion of the email goes where so we can identify the portion potentially responsible for promotional filtering;
• reworking the pieces of content going to promotion — whether it’s wording, images, or a link — to try to arrange something decent that still makes it to the main inbox;
• subsequently rebuilding the template as closely as possible to the original one while controlling that it goes to the main inbox.

In this case, we frequently use tools such as GlockApps to perform a substantial number of tests quickly.

S: Where do you see the sweet spot with the rise of AI-driven content and hyper-personalization? 

P: IMO, it’s already happening. We see a lot of AI-generated dynamic emails based on the product’s user experience and journey. Companies such as Kiliba in France provide 100% fully AI-automated management of the email system, triggering workflows and content based on intricate user behavior. My belief is that this will be more prevalent, allowing marketers to focus their attention on things such as A/B testing and cohort analysis.

S: How can brands automate email flows while maintaining strong engagement and avoiding spam filters?

P: Well, this is already happening as well, right? Automations automate your email journey. If you set them up and monitor them well, you should be able to maintain deliverability, like many brands currently do.

Wrapping up

We are incredibly grateful to Pierre Pignault for sharing his knowledge and experience. Here are some insights from this interview:

1. Waterfall segmentation improves engagement and deliverability — sending frequency should align with subscriber activity levels to maintain a healthy sender score.
2. Subdomain segmentation is a must — it protects your main domain reputation and helps monitor issues more precisely through tools such as Google Postmaster.
3. Avoid blindly resending to non-openers — it can damage your reputation. Instead, use automated re-engagement flows with well-defined entry rules.
4. The promotions tab isn’t the enemy — it can actually reduce spam complaints, and strategic content testing helps improve placement without sacrificing design.