Hey, it's Chase and Jimmy here.

A lot of brands have turned on AI features in their email platform, but very few are actually using them well.

They expect AI to magically improve performance without cleaning their data, testing recommendations, or thinking about how it fits into their strategy. Then they're surprised when results don't change or, worse, when things get more complicated.

AI can genuinely improve your email program (faster personalization, smarter segmentation, better optimization) but only if you implement it intentionally instead of just flipping switches and hoping for the best.

Today we're walking through the best practices that actually make AI work in email marketing, from clean data to smart testing.

Also inside:
✔️ In a world of ChatGPT copy, here's how to sound like you.
✔️ Inside the send: How Polish Pops nailed their spring collection launch
✔️ Quick clips: This week's top eCom news stories

Let’s dive in.

In a world of ChatGPT copy, here's how to sound like you.

AI makes content creation faster... but at what cost? When everyone sounds "friendly and approachable," nobody stands out. Join Omnisend on March 26 at 7:00 AM to build a brand voice that's actually recognizable across every email, ad, and social post.

What you'll learn:

• How to define your brand voice in 3-5 clear traits (not vague adjectives like "authentic")

• The difference between voice (what never changes) and tone (what adapts by situation)

• How to keep your messaging human and consistent without sounding robotic

Walk away with actionable guidelines your team (and your AI tools) can use without turning every message into the same beige copy. → Register now

Best practices for using AI in email marketing

AI has real potential to improve your email program, but most brands are either using it wrong or expecting it to do things it can't.

The difference between AI that drives results and AI that just adds complexity comes down to how intentionally you implement it.

You can't just turn on AI features and expect better performance. You need clean data, clear goals, and a plan for how AI fits into your broader email strategy.

Here's how to actually use AI in email marketing so it improves results instead of just creating more work.

1. Start with clean, organized data

AI is only as good as the data you feed it, which means messy or incomplete data will lead to inaccurate recommendations and poor performance.

If your customer data is inconsistent, full of duplicates, or missing key behavioral signals, AI can't make smart decisions on your behalf.

What to do:

• Clean your email list regularly by removing duplicates, invalid addresses, and unengaged subscribers who haven't opened in 90+ days

• Tag and segment subscribers accurately so AI has clear signals about preferences, behaviors, and lifecycle stages

• Track key behaviors like purchases, browsing, cart abandonment, and engagement consistently across all touchpoints so patterns are reliable

2. Use AI for personalization, not just automation

Automation saves time and reduces manual effort, but personalization is what actually drives conversions and builds customer relationships.

AI should help you send more relevant, timely emails that feel tailored to individual subscribers, not just help you send more emails faster.

What to do:

• Personalize subject lines, product recommendations, and content blocks based on actual behavior like browsing history, past purchases, and engagement patterns

• Use dynamic content blocks to tailor different sections of the same email to different segments without creating multiple versions manually

• Test AI-generated personalization against your standard approach to measure incremental lift and validate that the added complexity is worth it

3. Let AI handle repetitive tasks, not strategic decisions

AI is excellent at execution and optimization based on patterns, but it's not equipped to make strategic decisions about what your email program should accomplish or how it fits into your broader marketing goals.

What AI should handle:

• Segmentation based on behavioral data and engagement patterns

• Send time optimization for individual subscribers

• A/B test analysis and winner selection based on statistical significance

• Content generation and variation testing (with human oversight and editing)

What you should handle:

• Campaign strategy, goals, and how email fits into the customer journey

• Brand voice, messaging hierarchy, and creative direction

• High-stakes decisions that require judgment, context, or cultural awareness

• Determining which metrics matter and what success looks like

4. Always review AI-generated content before sending

AI can write decent, functional copy that checks basic boxes, but it can't write copy that sounds authentically like you or captures your brand's unique personality.

What to do:

• Use AI to generate first drafts or multiple variations that give you options and save time

• Edit thoroughly for tone, voice, accuracy, and brand consistency so the final version actually sounds human

• Check for awkward phrasing, factual errors, generic language, or messaging that doesn't align with your positioning

• Never send AI-generated content without human review, even if it seems good enough at first glance

Pro tip: AI works best as a brainstorming partner and efficiency tool, not as a replacement writer that eliminates the need for human creativity and judgment.

5. Test AI recommendations instead of blindly following them

AI suggestions are based on patterns and probabilities, not guarantees, which means they should be validated through testing before you rely on them completely.

What to do:

• A/B test AI-generated subject lines against your own to see which approach actually performs better with your audience

• Compare AI-optimized send times to your standard schedule to measure whether the added complexity improves results

• Track performance of AI-driven segments versus manual segments to validate that the automation is making smarter decisions

• Use data to validate or challenge AI recommendations instead of assuming they're always right just because they're generated by an algorithm

6. Use AI to improve deliverability and inbox placement

AI can help you avoid spam filters, improve inbox placement, and maintain strong sender reputation by identifying issues before they hurt performance.

How it works:

• AI-powered tools analyze your email content for spam triggers like overuse of promotional language, suspicious links, or formatting issues

• They suggest improvements to subject lines, sender names, and body copy that reduce the likelihood of being flagged

• Some tools verify email addresses in real time to reduce bounces and protect your sender reputation

Pro tip: Pair AI deliverability tools with strong list hygiene practices like regular suppression of unengaged subscribers and double opt-in for best results.

Real examples of AI in email marketing

AI-generated subject lines

Tools like Omnisend, Klaviyo, and other email platforms offer AI subject line generators that analyze your campaign goals and past performance to suggest options with higher predicted open rates.

How to use it:

• Input your campaign keywords, a brief description of the content, and any specific goals like urgency or curiosity

• Review AI-generated options and select the ones that align with your brand voice and campaign strategy

• A/B test AI suggestions against your own subject lines to see which approach actually performs better with your specific audience

• Track which approach consistently wins over time to learn whether AI is improving your results or just adding complexity

Pro tip: Use AI to generate subject line variations you wouldn't naturally think of, then edit them for brand voice and test them against your instincts to expand your creative range.

AI-powered product recommendations

AI analyzes browsing and purchase behavior to suggest relevant products in emails, increasing the likelihood that subscribers will find something they actually want.

Where this works:

• Post-purchase follow-ups that suggest complementary products based on what someone just bought

• Browse abandonment emails that show items similar to what someone viewed but didn't purchase

• "You might also like" sections in promotional campaigns that surface products based on past behavior

Pro tip: Combine AI recommendations with manual curation for best results by letting AI surface relevant options based on data, then guiding the final selection based on strategy, inventory, or promotional goals.

Predictive send time optimization

AI tracks when individual subscribers typically open and click based on their historical behavior, then schedules emails to arrive during those optimal engagement windows.

Where this works:

• Welcome flows where timing flexibility matters less than individual optimization

• Nurture sequences and educational content where engagement is more important than coordinated timing

• Ongoing campaigns to engaged subscribers who have established patterns

Where it doesn't work as well:

• Time-sensitive campaigns like flash sales or limited product drops where coordinated timing matters more than individual optimization

• Multi-channel launches where email needs to align with other marketing activities happening at specific times

Dynamic content personalization

AI swaps in different content blocks, images, or messaging based on subscriber behavior, preferences, or segment membership without requiring you to build separate emails for each group.

Examples:

• Showing different hero images based on past purchases or product category preferences

• Recommending products based on browsing history or items left in cart

• Tailoring copy and messaging based on lifecycle stage, engagement level, or customer value

Pro tip: Start with one dynamic element like product recommendations or category-specific imagery before personalizing entire emails, so you can measure impact and avoid overcomplicating your campaigns.

Implementation over features

The brands getting real results from AI aren't the ones using every feature available. They're the ones who identified specific problems in their email program and used AI to solve them.

Start by asking where AI can actually help. Are you spending hours manually segmenting lists? Struggling to personalize at scale? Guessing at the best send times? Those are the places where AI delivers measurable impact.

But don't implement AI just because it's available. Every feature you turn on adds complexity, requires clean data, and needs ongoing management. Focus on the capabilities that solve real problems and improve results you actually care about.

Test everything. Just because AI suggests something doesn't mean it's right for your audience. Validate recommendations with A/B tests and track whether AI-driven decisions are actually improving performance over time.

Most importantly, remember that AI is a tool for better execution, not a replacement for strategy. The brands winning with AI are using it to do what they already do well, just faster and at greater scale. That's the difference between AI that drives growth and AI that just creates more work.

Inside the send: How Polish Pops nailed their spring collection launch

Inboox's AI Breakdown analyzed this seasonal drop email and found what popped:

• Colorful, on-brand design that screams "spring" without saying a word

• Prominent CTA button that makes the next step crystal clear

• Clean layout with social proof touchpoints that build credibility fast

It also caught low-hanging fruit like more product imagery to fuel FOMO, a higher-contrast CTA to boost clicks, and a personalized greeting to make subscribers feel seen (not just sold to).

That's the power of Inboox. Access 1.5M+ real Shopify emails, get instant AI breakdowns on what works (and what doesn't), plus send-time data, full HTML, and more.

→ Start your free trial at Inboox.ai

Quick Clips:

• A match made in (misfit) heaven: ThredUp and Misfits Market just linked up so you can ship your old clothes and get grocery credit in return. Two brands built around rescuing "rejects" found their match.

• Clicks and bricks, still the mix: Forrester projects U.S. eCom hitting $1.8 trillion by 2030, but stores will still own 71% of retail sales. Physical retail hasn't just survived... it's become the experience layer online can't replicate. The takeaway? Omnichannel isn't a buzzword, it's still the play.

• AI day is the new Sunday reset: At NRF 2026, Emma Grede, Ben Francis, and Shopify's Harley Finkelstein went off-script on where retail's headed. Shopify merchants saw a 14x jump in orders from agentic apps in one year. Grede now blocks off a weekly "AI day" and treats it like a second executive brain, and Francis says Gymshark's secret is staying focused on what they're best at and letting partners handle the rest.

Annnnd that’s a wrap for this edition! 

Thanks for hanging with Chase and me. Always a pleasure to have you here.

If you found this newsletter helpful (or even just a little fun), don’t keep it to yourself! Share ecomemailmarketer.com with your favorite DTC marketer. Let’s get them on board so they don’t miss next week’s drops.

Remember: Do shit you love.

🤘 Jimmy Kim & Chase Dimond

PS - Your next best customer might be reading this right now. Want in? Email Jimmy to sponsor this newsletter and more.

Love this newsletter but want to receive it less frequently? Let us know by clicking here!