Users go quiet after the first week
Don't send generic "come back" emails. Show users what they're missing by including their actual data and progress.
When to use
You have users who signed up and used the product but went inactive. You have their usage data.
Hypothesis template
If we send personalized re-engagement emails showing [user's data/progress] after [X] days inactive, we'll reactivate [target]% of churned users.
Method
The problem: Users go inactive and generic "we miss you" emails get ignored. They feel like spam.
What works: Emails that include the user's actual data. "Your project has 3 incomplete experiments" or "You were 60% through setup — here's what's next."
The result: Personalized re-engagement emails get 3-5x higher click rates than generic ones because they remind users of their investment.
How to do it:
- Define "inactive": no login for 7 days? 14 days? Pick based on your product's natural usage frequency
- Build an email that pulls in the user's actual data: progress, incomplete items, recent activity
- Include one clear action: "Finish your experiment" or "Check your results"
- Send at day 3, day 7, and day 14 — then stop (3 emails max)
- Make the CTA go directly to where they left off, not the homepage
- Track reactivation rate per email
Key insight: "You have 3 unfinished experiments" works 5x better than "We miss you!" because it reminds them of sunk cost.
Success metrics
- •Email open rate
- •Click-through rate
- •Reactivation rate (returned within 7 days)
- •Long-term retention of reactivated users
Prerequisites
- User activity tracking to know when someone goes inactive
- Email system that can include dynamic user data
- Deep links to specific in-app locations
Common pitfalls
- •Sending generic "we miss you" without personalization
- •Sending too many emails (3 max in the sequence)
- •Linking to the homepage instead of where they left off
- •Not segmenting by how far they got before going inactive
Source: Common retention pattern. Personalized data in re-engagement emails outperforms generic messages 3-5x.
Suggested ICE scores
Ready to run this experiment?
Sign up free to use this playbook with step-by-step guidance and track your results.