Writing for two readers: How email marketers should adapt their strategy as AI enters the inbox
You’ve probably seen it already. AI writes an email in seconds, but the copy feels generic. AI helps you produce more emails, but engagement still drops. Every new AI tool promises to automate your workflow, and it becomes harder to tell which of them actually improves results.
The challenge is no longer whether to use AI. It’s about using it without hurting engagement or your brand voice. That’s because speed is easy to automate, while strategy isn’t. AI can generate ideas and draft copy, but it doesn’t understand your audience, your business priorities, or the context behind your decisions.
We discussed these aspects in a webinar with Des Brown and Dmytro Kudrenko. Dmytro Kudrenko is the founder & CEO of Stripo. Des Brown is an email marketing strategist with 13+ years of hands-on experience in email and digital, the founder of Email Expert Africa, Email Advice in Your Inbox, and CrossLetter, and a proud Stripo ambassador.
In this article, you’ll learn where AI really improves email marketing, where it still depends on human expertise, and what you should change in your email workflow to stay effective as AI reshapes the inbox.
You can watch the webinar on the Stripo YouTube channel:
Key takeaways
- AI delivers the most value in email marketing when it helps you draft content faster, personalize campaigns using customer data, and run tests at scale.
- Your emails now have two readers: your subscribers and AI. Optimize for both with semantic structure, descriptive alt text, and text-based CTAs.
- Some things haven’t changed with AI, and they remain very important: List hygiene, authentication, engagement, consistent sending, and welcome emails matter just as much in the AI era.
- Don’t rely on open rates alone. Focus on metrics that reflect real business outcomes, such as purchases and subscriptions.
- To make AI as useful as possible for your content, give it authentic examples of your own writing. Create folders with context about your company, and “send” AI to those folders before creating a draft. Always review and edit AI-generated texts to avoid generalizations and false statements.
Where AI actually helps and where it breaks
The biggest win AI brought to email marketing in 2025 was speed. According to Litmus, 62% of teams said in 2024 that producing a single email took them two weeks or more. By 2025, that number had dropped to just 6%.
AI made content production much faster, but it did not replace strategy. It is a tool that amplifies your strategy rather than replacing it.
That’s why marketers need to understand where AI creates real value and choose the right tools for those tasks.
(Source: Des Brown presentation)
Content and copy at speed
AI beats humans when it comes to drafting content quickly. It will generate subject line variations, preview text, and body copy drafts faster if you provide the right context and clear instructions.
But as with everything that makes a process easier, there’s a catch. And if the input is weak, the output is going to be weak.
So if your prompt is generic or you don’t have enough data, or the essence of the email is unclear, what you get from AI will be just as shallow and formulaic. Faster, yes, but is it of higher quality?
Most people in email are pretending really well. AI hasnʼt changed that…
itʼs just given them a better way to do it.
Personalization signals
AI is really good at things like pattern recognition. AI helps take that and speeds it up by producing dynamic content, triggering messages based on behavior, and optimizing send times for when people are actually in their inbox.
But this only works if you already have enough customer data and well-segmented email lists. AI can’t come up with personalization out of thin air. Technically, it can, but it will do more harm than good.
You need solid data to build on what AI can use to help you within the email environment.
Des Brown,
Testing at scale
Another area where AI clearly and unquestionably helps email marketers is testing. Experiments that used to take months to complete now take days, with results available almost immediately.
The testing opportunities are virtually endless: subject lines, CTAs, email layouts, and other campaign elements that you have doubts or, conversely, hypotheses about.
Where it breaks: Polished content sounds generic
This is where things become tricky. If you don’t have a strategy, if you just mindlessly use AI for everything, your content starts to feel generic and predictable. Because content is about meaning and unique perspective, the thoughts that are going through your head, the problems you’re thinking about, and the solutions you’re coming up with.
AI is designed to guess the right sequence of words based on combinations it’s seen before. So AI-generated writing feels incomplete and lacks life and soul.
Polished but generic content is starting to cost senders.
Des Brown,
If you’re a brand or a newsletter sender, people are signing up for a specific reason, and that’s to hear from you. Inboxes are also changing, and standing out is getting a lot tougher. If you’re trading out generic content, you often wind up getting lost in that noise.
The fix isn’t to avoid AI but to use it in the service of a distinctive voice.
So AI, as a framework and as a tool to support what you’re doing in the email space, becomes really valuable because it buys you time. But if there’s no distinctive voice, if there’s no human behind what’s being sent, that is where the problems start creeping in.
Des Brown,
How inbox providers respond: AI detection and deprioritization
AI is changing email marketing not only from the sender’s side. Inbox providers are also using AI to their advantage, and in many cases, that works against senders.
- Inbox providers are building AI-detection signals.
Generic, same-sounding content earns fewer opens and clicks. Inbox providers perceive these weak engagement signals as indicators of low-value or AI-generated content and deprioritize these messages in the inbox. AI-generated summaries make this even more apparent: If recipients rely on the summary to decide whether to open an email and choose not to engage, senders lose the interactions that help maintain strong inbox placement. - Emails that sound the same get disengaged in the background.
Email senders are no longer getting the opens they used to get, the clicks they hoped to attract, and the actions they wanted people to take in those emails. When brands rely on generic AI-generated copy, they make it less likely that subscribers will engage with their emails. Lower engagement then reinforces the negative signals inbox providers use to evaluate future campaigns. - AI-assisted cold email at scale is poisoning shared infrastructure.
When a sales team uses AI to send cold emails at scale, spam complaints can damage the reputation of the sending domain, and that becomes a problem for the entire company. Once domain reputation declines, all emails sent from that domain will suffer. Even legitimate messages that recipients are waiting for may end up in the spam folder. It takes time for the sender’s reputation to recover.
You’re now writing for two readers
As AI-powered inboxes become more common, AI has effectively become another participant in email communication. Marketers are now writing for two readers, which means they need to optimize emails for both people and AI.
AI opens an email and generates a summary before a subscriber even reads it. As we discussed earlier, engagement signals such as opens, clicks, and reading time continue to influence deliverability and inbox placement. That is why senders are interested in structuring their emails so that AI can easily understand them. The goal is not just to help AI scan the content but to make sure it recognizes the value your email provides.
The good news is that the same best practices benefit everyone. So, you need to stick to the basics that work equally well for human readers, AI, and accessibility, which is also important now. In other words, what makes an email accessible also makes it easier for AI to understand.
A checklist for making emails legible to AI:
- Use semantic HTML and a clear email structure. Organize your content logically so AI can easily understand the flow of the message and identify where to find the call to action.
- Use a semantic heading hierarchy and descriptive headings. Clear headings help both readers and AI understand the purpose of each section and navigate the email more efficiently.
- Write meaningful and solid alt text for every image. Descriptive alt text makes your content understandable for all recipients, including people using assistive technologies and AI. Labels such as “image1” or “banner” provide no useful context, making it harder for AI and subscribers to determine whether the email is worth reading.
- Don’t hide your call to action within an image. AI cannot recognize or act on a call to action if it is presented solely as an image. Text-based CTAs are easier for both AI and subscribers to understand.
If these basic rules aren’t followed, AI has less information to work with when generating an accurate summary and evaluating the relevance of your email.
Accessibility is no longer just about people. It also helps AI understand the value your emails provide.
What to test now that open rate is broken
Open rate is no longer a reliable metric. AI-powered inbox features increasingly open emails to generate summaries or preview content. As a result, an open no longer guarantees that a person actually read your email. What to test, then, and how to measure whether emails work as they should?
Instead, focus on metrics that reflect real engagement and business outcomes. Test clicks and conversions to understand what your audience is actually telling you. Look for patterns in what drives engagement, what falls flat, and which campaigns make you profitable.
It is also worth testing your email subscriber acquisition cost. If you have paid acquisition channels and track the customer journey from the first touchpoint to becoming a paying customer, measure the performance of the channels where that journey begins. It’s a good way to find out what people are doing along the way and how it can be made faster or better.
The next thing worth testing is your email design. Experiment with different layouts, content structures, visual hierarchies, and CTA placements. As you collect more data, you’ll better see what people pay more attention to, where they click, and which layouts work best. All of this works and helps you make better decisions.
Now you also need to test how AI interprets your emails. As AI-generated summaries become more common, the structure of your email can influence how your content is presented.
Depending on how you structure your content, the headings you write, and what you put underneath them, the quality and accuracy of the summary AI generates will also change. You need to collect enough data to know how to adapt your emails to the AI.
There are two ways to get reliable testing results. The first is to have a large audience. The second is to send a consistent volume of emails over time.
For example, testing a subject line in a single campaign and concluding that it works is not enough. If you compare subject lines with and without emojis over a period of six months, you can have more confidence in the results.
Important factors that affect this are the type of sender you are and the volume of email you send. The more consistent your sending patterns and the larger your dataset, the more accurate your conclusions will be.
What hasn’t changed and why it matters more now
While many things change with the use of AI in email marketing, some fundamentals remain the same. No matter how advanced the tools become, these core principles continue to have the greatest impact on email marketing success.
- List health
If you don’t put your readers first, if they aren’t interested in the content you send, your deliverability will suffer. Even the most polished AI-generated emails cannot compensate for a disengaged audience. - Authentication
You need to set up your technical environment correctly. Authentication protocols verify your identity as a sender and prove that your domain can be trusted.
AI can help you optimize your campaigns, but it can’t fix a weak technical foundation. If your authentication is not configured correctly, even the best AI-generated emails will struggle to reach the inbox. - Engagement signals
All these signals, such as opens, clicks, and replies, already affected deliverability. Inbox providers continue to rely on these metrics when deciding whether an email belongs in the inbox or the spam folder. That hasn’t changed.
What’s new is that the same signals also influence how emails are prioritized within the inbox. They can affect whether AI highlights your message or pushes it further down the list.
There is no need to optimize separately for AI. AI evaluates the same engagement behavior your subscribers already give you. - Sending behavior
If your email strategy is consistent and your campaigns are sent on a predictable schedule, you are in a better position. The inbox is being curated around trust, and trust is driven by consistency.
If you want your emails to stand out and increase their chances of being prioritized by AI-powered inbox features and summaries, you need to send consistently. And that is good email practice across the board. Good for deliverability and good for the relationships with your subscribers you’re building. - The welcome sequence
This is still one of the highest leverage automation email sequences you can build. And it’s easier to waste this window with AI in place.
Many marketers make the wrong choice when it comes to welcome emails. They want to take a shortcut: make welcome emails templated and generic to get people in quickly and then send them ads or purchase offers later. Don’t do that.
A welcome email is one of the highest performing messages in the customer journey. According to Invesp, welcome emails generate 4 times more opens and 5 times more clicks than regular email marketing campaigns. Make it valuable enough to encourage opens, clicks, and other positive engagement signals from the very beginning.
Even if you use AI to create your campaigns, your welcome sequence deserves extra attention because the value of that welcome sequence hasn’t changed at all.
Here are three principles to keep in mind because they will remain true no matter how AI reshapes email marketing:
- Build your list with intent. Focus on attracting people who are truly interested in your business or products. Reaching the right audience is the foundation of an email strategy that supports long-term business growth.
- Deliver value to your subscribers. You need to earn a place in the inbox. No matter who technically “runs” the inbox or how AI changes the email experience, your messages will remain a priority if your subscribers value what you’re sending.
- Respect your readers. Avoid sending mass emails that are too broad or irrelevant. The more relevant and thoughtful your communication is, the stronger your subscriber relationships will become, and the better your long-term email performance will be.
Make sure you do the basics well, because they won’t change regardless of what happens with AI in the inbox.
Three questions before you add any AI tool to your stack
Before you adopt any AI tool in your email stack, you need to ask yourself these questions:
- Does it serve my strategy?
There is a danger in starting to use an AI tool that promises to write an email in under two minutes while matching your brand voice and tone. The risk is that marketers will send AI-generated content with little or no review, assuming the output is ready to go.
Before adopting any AI tool, ask whether it supports your strategy or replaces it. Choose AI tools that make your strategy stronger, not just faster. If a tool helps you create better campaigns while staying aligned with your goals, it’s worth adding to your workflow. - Will it affect my deliverability?
If your emails never reach the inbox, nothing else matters. That’s why deliverability should be one of the first things you evaluate before adding any AI tool to your workflow.
Be careful with tools that make bold promises but haven’t been thoroughly tested. Some AI-powered solutions are built quickly and focus more on marketing claims than proven results. If a tool negatively affects your ability to reach the inbox, it can undermine every part of your email marketing strategy. - Can I measure its impact?
It’s easy to get caught up in the excitement around the latest AI trends. New tools appear every week, and it’s tempting to try them all.
The more important question is whether you can measure the impact of a tool. Can you prove that it speeds up your workflow, helps you scale your email marketing efforts, or reduces the time required to complete tasks?
If you don’t have the answer or can’t measure this impact, you’re paying a monthly fee for a tool that adds little or no value. Every AI tool should contribute to a stronger strategy or a better product. If its impact can’t be measured, it’s worth reconsidering whether it belongs in your stack.
(Source: Des Brown presentation)
What’s important to remember is that the tools are changing, but the job of an email marketer hasn’t changed. You still need to focus on what your audience values because engagement signals continue to matter. Ignoring AI is not an option because it’s not going away. And your competitors will use AI to test campaigns faster, analyze data more effectively, and learn from their results more quickly.
AI is changing email marketing, but it’s not doing your job for you. It simply makes your strengths and weaknesses more visible.
Using AI comes with risks. This applies to both the delivery and the management of what your AI tool produces. Fortunately, many of them can be minimized with the right workflow.
With Stripo, you can create a Brand Guidelines kit so the AI Assistant generates emails using only your approved colors, fonts, and other brand assets. This helps keep every email consistent with your brand identity.
You can also use synchronized modules and instruct AI to build emails only from preapproved content blocks. Instead of generating layouts from scratch, AI works within components your team has already reviewed and approved. Stripo can also help you cut email production time by over 3.7 times.
Approval workflows and an email design system add another layer of quality control. They help catch mistakes before an email is sent and ensure that every campaign meets your brand and compliance standards.
Automation should not introduce unnecessary risk. It should be predictable and controlled. And this can be provided by a governed builder, which is Stripo.
The workflow: How to teach AI your voice
If you write a prompt and ask AI to create some content, the result will likely look quite templated and generic, even if you add a few examples of your previous articles or social media posts beforehand.
Several things affect the quality of a result:
- review and edit every AI-generated draft. Human input is what turns a generic draft into content that feels accurate, credible, and aligned with your goals;
- teach AI your tone of voice, brand parameters, and context. Provide clear guidance on your tone of voice, brand values, audience, and writing style. The more context AI has, the more relevant its output becomes;
- use AI for the framework and your expertise for the brand voice. Let AI handle the structure and first draft while you shape the messaging, personality, and perspective that make the content unmistakably yours.
Build a framework that teaches AI your brand voice, messaging, and context so it can generate good first drafts. But never skip the human review. Make sure every draft is reviewed by someone who truly understands your brand.
As we mentioned earlier, weak prompts lead to weak results. You have to provide everything you know about your brand, store it in the AI, and reuse it so you don’t need to keep submitting it.
Here is a practical workflow for teaching AI your brand voice.
- Organize context folders. Create a dedicated folder to store everything related to your brand and its tone of voice. Include information about your company, goals, product, positioning, audience, and messaging. Send AI to this folder every time you prepare content so you don’t have to rebuild the context from scratch.
- Build a voice markdown file. Record yourself talking about your brand for an hour. Explain your positioning, values, strengths, target audience, and the way you communicate. Turn that recording into a markdown document that AI can use as a reference.
- Share your existing content. Give AI 100 to 200 of your LinkedIn posts or other pieces of content you’ve written. This helps it recognize your writing style, structure, vocabulary, and format and the topics you typically cover.
- Keep your context up to date. Add new articles, posts, and other relevant materials to your brand folder as they’re created. A constantly updated knowledge base helps AI reflect your latest messaging instead of repeating the same ideas over and over again.
- Avoid feeding AI with AI-generated content. Train it using content you wrote yourself, without AI polishing or rewriting. AI needs authentic examples of your writing to learn your voice. If you feed it content that was generated by AI, the output will become even more generic.
- Review every AI-generated draft. Never publish content without checking it first. You never know when or why it might hallucinate and invent something that isn’t there, or approach the topic from a different angle than you imagined. Human review is always essential.
- Keep your knowledge base organized. The AI relies on your organizational skills, and clear folders, well-named documents, and specific instructions about where to find information reduce prompting time and produce more consistent results.
- Build a prompt library. Save prompts that produce good results and reuse them in future projects. Keep the prompt itself consistent, but update the supporting context for each new task.
You use AI to take data and make sense of it, and you make decisions based on that data. Don’t let AI make that decision for you.
Des Brown,
How to choose an AI tool?
There is no universal answer here. The best tool is the one that fits your workflow and gives you enough control over context, brand voice, and outputs. For example, you can use Claude to build reusable frameworks, organize your brand knowledge, and store the context AI needs to generate better drafts over time.
To choose the right AI tool, you can focus on the following criteria:
- Does it give you control over context? Can you upload knowledge bases, brand documentation, writing samples, etc.?
- Can you measure its impact on your workflow? Look for evidence that it saves time, improves quality, or increases productivity.
- Does it integrate with your ESP or CRM? The more context AI has access to, the more useful its recommendations and outputs will be.
If your email service provider supports it, you can also connect it to AI through MCP (Model Context Protocol). MCP lets an AI assistant connect directly to tools your ESP provides, giving it access to the data and context it needs to provide more relevant insights and recommendations, rather than relying solely on the information you include in a prompt.
There are widgets in Stripo; it’s a really fantastic way to use AI to prompt and create dynamic or interactive content in emails.
Don’t focus too much on tools. You can always change them. The only thing that you need to think about is human. I say individuals and interactions over processes and tools.
Dmytro Kudrenko,
(Source: Des Brown presentation)
In Des Brown’s view, the next stage of AI in email marketing will move beyond automation and toward true one-to-one communication.
The goal is to make every email perfectly tailored to each individual recipient. Instead of sending one campaign to thousands of subscribers, marketers will send thousands of unique emails, each adapted to a specific person.
That future depends on having richer customer data and tighter integration between your email design environment, your ESP, and your audience data. With enough context, AI could understand what each subscriber is interested in and what motivates them to engage and adapt the message for each recipient.
That is where email marketing is heading. “How can AI help us create emails faster?” is now the wrong question. You need to ask: How can AI help us create a better email experience for every subscriber? How can we move from sending one email to thousands of people to sending an email of one? And how can we create a truly individualized email experience for every recipient?
Wrapping up
AI is now a part of life for both consumers and email marketers. Ignoring it or being overly cautious about using it means missing opportunities for your business. Email marketers need to understand where AI creates the most value and build workflows that keep decision-making and strategic thinking in human hands.
Whether you’re an experienced AI user or a newbie, take one practical step after reading this article. Identify where AI delivers the greatest value, then ask yourself whether you can actually measure its impact. Before adopting another AI tool, come back to the three questions Des Brown recommends asking before adding it to your workflow.
Many thanks to Des Brown for sharing his insights on the role of AI in email marketing and his practical advice on teaching AI to write in your brand voice.
FAQ
1. Will using AI to write my emails hurt my deliverability?
It’s not that using AI for writing emails can hurt your deliverability. Your deliverability will be hurt if the quality of the content AI produces is low and your subscribers don’t engage with your emails as a result.
In other words, generic content leads to lower engagement, and lower engagement in turn weakens deliverability. The problem is not AI itself. The problem is publishing AI-generated content that isn’t relevant or useful.
2. If open rate is unreliable now, which metrics should I actually test?
Focus on metrics that reflect engagement and business outcomes. Pay close attention to clicks, conversions, and the actions of subscribers after they open your email.
Measure how email performance contributes to real business results: purchases, signups, or downloads.
3. How do I stop AI tools from making all my emails sound the same?
Teach the AI your tone of voice, give it lots of examples of your writing, always review and edit the drafts the AI has prepared, and reuse prompts that have produced strong results.
4. What should a human always review before an email goes out?
Review the content to make sure it reflects your brand voice and that every claim is accurate. A human should also make the final decisions based on the data. AI can analyze patterns and suggest recommendations, but it shouldn’t decide your strategy.
Treat AI like a junior assistant: Delegate tasks, but review everything before it goes out under your name.
0 comments