I need to tell you something that’s been bothering me for months now, something most marketing experts won’t admit: we’re teaching people to hate personalized emails.
Every morning, I watch my inbox fill with messages that start with “Hey Sarah!” (not my name, but you get it). They reference products I looked at last Tuesday. They arrive at exactly 9:47 AM because some algorithm determined that’s when I’m most likely to click. And you know what? I can smell the automation from a mile away.
So can your customers.
While every marketing conference celebrates AI’s power to personalize subject lines and optimize send times, we’re missing something critical: when personalization becomes ubiquitous, it stops feeling personal.
The Problem Everyone’s Ignoring
Here’s what’s happening. We’ve created what I call “intimacy inflation”-where signals of personal attention have become so automated and abundant that they now signal the exact opposite of what we intended.
Think about your own inbox. When you get an email that:
- Includes your name in the subject line
- Recommends products “just for you”
- References your browsing history
- Arrives at your “optimal engagement time”
…do you think “Wow, this brand really understands me”? Or do you think “I know exactly which marketing automation platform sent this”?
Be honest.
We’ve industrialized intimacy, and people see right through it. Every brand sends abandoned cart emails now. Every company has birthday promotions. Everyone’s doing behavioral triggers. The competitive advantage isn’t doing these things better-it’s doing something completely different.
What AI Gets Wrong About People
The biggest flaw in AI email personalization isn’t technical. It’s philosophical. We’ve confused pattern recognition with understanding.
When Less Is Actually More
The most sophisticated brands I’ve worked with aren’t maximizing personalization-they’re being strategic about when to pull back.
At Sagum, we tested something counterintuitive: mixing highly personalized sequences with deliberately un-personalized “brand perspective” emails. Same message to everyone. No name tokens. No behavioral triggers. Just compelling ideas and honest communication.
The results surprised us. Those un-personalized emails often outperformed our “optimized” sequences. Why? Because the contrast made them feel more human. The unpredictability felt like an actual conversation instead of a marketing funnel.
The Context Problem
Here’s what AI can’t see: AI analyzes historical patterns brilliantly, but it’s terrible at understanding what’s happening right now.
Your algorithm knows I usually open productivity emails on Tuesday mornings at 10 AM. What it doesn’t know is that this Tuesday, I’m dealing with a family emergency. Or celebrating a major win. Or nursing a hangover. Your “10 Productivity Hacks” email that would normally land perfectly now feels completely tone-deaf.
The smartest email strategies don’t just personalize to the individual-they personalize to the moment that individual is experiencing. That means incorporating cultural context, economic signals, seasonal patterns, and yes, even weather and current events into your strategy.
You’re Optimizing Away Discovery
Here’s something that keeps me up at night: when your AI gets really good at predicting what each subscriber wants, you trap them in a filter bubble.
They never see:
- Products they didn’t know existed
- Ideas that challenge their assumptions
- Content that could deepen their relationship with your brand
You optimize your way into irrelevance.
The solution? Build what I call “strategic serendipity” into your email program. Deliberately calculated moments of randomness that break the predictable pattern and create genuine discovery.
How to Actually Fix This
Look, I’m not suggesting you abandon AI. That would be stupid. But I am suggesting you completely reframe how you use it.
Stop using AI to write your emails. Start using it to make your human writers smarter.
AI as Intelligence, Not Author
Here’s a better model: use AI to analyze your subscribers and identify distinct mindsets-not demographics, but actual behavioral and preference patterns. Maybe you discover 12 different “types” of subscribers based on how they engage with your content.
Then have actual humans craft email tracks for each mindset. Not with mail-merge personalization, but with fundamentally different value propositions, tones, and content strategies. Your AI’s job is matching people to the right track, not generating synthetic personal touches.
Be Honest About the Automation
The boldest move? Stop hiding the fact that you’re using AI.
“Our system noticed you’ve been researching…” is actually more trustworthy than pretending a human personally reviewed each subscriber’s behavior. Transparency about automation paradoxically increases authenticity.
People aren’t stupid. They know you’re not personally thinking about them. Stop pretending you are.
Build Personalities, Not Tokens
Stop thinking about personalization as inserting [FIRST_NAME] or showing [RECOMMENDED_PRODUCT]. Instead, create distinct email personalities-complete editorial voices with different:
- Tones (irreverent vs. authoritative vs. confessional)
- Depths (quick-hit vs. deep-dive)
- Balances (educational vs. promotional)
Your AI matches subscribers to personalities. The personality does the heavy lifting.
Deliberately Limit Your AI
This sounds backwards, but it works: constrain what your AI can optimize.
Don’t let it personalize:
- More than 40% of your subject lines
- Send times more precisely than 4-hour windows
- More than 3 content blocks per email
Constraints force better strategic decisions and preserve the human fingerprint that makes your emails distinctive.
The Battle You’re Not Watching
Here’s something almost nobody talks about: while you’re optimizing your sender-side AI, Gmail and Apple Mail are deploying recipient-side AI specifically designed to protect users from marketing manipulation.
You’re in an arms race. Your AI tries to maximize opens and clicks. Their AI tries to minimize interruption and noise. Your subscriber is caught in the crossfire.
The only winning move? Build email programs that recipient-side AI wants to surface because they genuinely provide value. That means:
- Higher content-to-commerce ratios
- Longer intervals between sends
- More subscriber control
- Actual utility beyond selling
Basically, stop trying to game the algorithm and start respecting people’s time.
Why Your Vendor Won’t Tell You This
After managing millions in ad spend across platforms-including over $2 million on TikTok alone-I’ve learned something important: follow the money.
Most AI email vendors make money by:
- Processing more data
- Enabling more sends
- Adding more complexity
Their economic interests are directly opposed to what’s best for your subscriber relationships. The most effective approach is often simpler: fewer, better emails. Less personalization, more personality. Lower volume, higher impact.
But that doesn’t align with how they get paid, so they’ll never recommend it.
What to Do Right Now
If you want to actually fix your email program instead of just optimizing a broken strategy, here’s where to start:
First 30 Days
- Audit your current personalization ratio. If more than 80% is algorithmic, you’ve crossed into synthetic territory
- Launch one deliberately un-personalized sequence focused on brand perspective instead of behavioral triggers
- Start tracking relationship metrics (replies, feedback, conversation) alongside conversion metrics
Next 60 Days
- Develop 3-5 distinct email personalities and build routing logic
- Implement constraints on your AI personalization
- Test transparent messaging that acknowledges your use of automation
By 90 Days
- Restructure around conversation metrics, not just conversion
- Incorporate contextual signals beyond individual behavior
- Build strategic serendipity mechanisms
The Uncomfortable Truth
The future of AI email isn’t more sophisticated algorithms generating more personalized content. It’s using AI to help humans have more authentic conversations at scale-and having the courage to sometimes choose less personalization for more genuine connection.
The brands winning this game aren’t those with the most advanced AI. They’re the ones who understand that technology is only as good as the strategy behind it. And sometimes the most personalized thing you can do is treat subscribers like intelligent people who can spot fake intimacy instantly.
Because they can. They absolutely can.
Your open rates are already telling you this truth. The question is whether you’re ready to listen.
At Sagum, we’ve built our reputation on challenging conventional wisdom and focusing on long-term growth over short-term optimization games. Our approach to AI and email reflects this: sophisticated simplicity, strategic restraint, and genuine relationships at scale. Not because it sounds good in a blog post, but because it’s the only thing that actually works.
