When does AI personalization stop feeling personal and start hurting deliverability?
Every outbound tool now claims it can write emails that sound handwritten. Scrape a LinkedIn profile, pull recent posts, mention a specific detail, and the email reads like a real person spent five minutes on it. Reply rates on individual campaigns can look great for a while. But inboxes across the internet are filling up with the exact same pattern of "I saw you posted about X" openers, and spam filters are getting better at catching bulk sends that mimic personal writing.
I ran outreach fully manual for a while before building any automation around it. No AI, no templates copied across contacts, just reading each person's profile and writing something specific to them. That got a reply rate around 25%, which is high for cold email by most benchmarks. When I started automating parts of it to scale past what one person can write by hand, the reply rate didn't hold. Not because the automation was bad, but because the moment personalization becomes systematic, it starts pattern-matching to every other automated email a prospect has already seen that week.
The tradeoff I keep running into is that manual personalization doesn't scale past a small list, but scaling AI personalization eventually makes every email start to look the same to both spam filters and to prospects who've seen the pattern before. It feels like the more "personal" outreach becomes at scale, the less it actually reads as personal to the person receiving it.
For founders and marketers who run outbound at any volume, where do you draw the line between using AI to personalize and just adding automated noise to someone's inbox? What's actually changed in how you approach cold outreach because of this?
Replies
The pattern matching problem is the real one, not the personalisation itself. A prospect does not need to consciously spot the template. They only need to have seen the shape of it once before, and the whole email reads as manufactured even when every detail in it is true.
What changed it for me was moving the personal part earlier than the email. If the only place a prospect specific detail shows up is the opening line, it reads as decoration and gets discarded with the rest of the noise. If it shows up in what you are actually offering, the timing, the specific problem you are pointing at, decoration is not possible, because the offer itself stops making sense for anyone else.
That does not solve the scale problem, it just moves the ceiling. Instead of how many emails one person can write by hand, the limit becomes how many genuinely different situations you can actually tell apart before you start grouping people who are not really alike and the automation starts showing.