Editorial banner showing different player segments returning, with only a few driving most of the real value.

Reactivation Looks Good Until You See Who Came Back

Reactivation campaigns can look healthy even when they hide where the real value sits.

The useful shift came when success stopped meaning "users came back" and started meaning the right players returned at the right cost and actually became active again.

In-Short

Positive Results Hid the Wrong Question

Campaigns showed growth in active users and quick payback. That looked like success. But top-line results did not show which players were actually worth bringing back.

Two Segments Carried Most of the Value

Once users were split by deposit behavior, a clear pattern appeared. Recently dropped players and strong historical depositors reacted best, while other segments looked weaker when isolated.

Budget Shifted Without Increasing Spend

The change was not about spending more. It was about spending smarter. Stronger segments received better boosts and more attention, while weaker ones stopped being treated equally.

Not All Players Restart the Same Way

Reactivation is not one action. Some players return easily. Others need effort and still bring little value. Once you see this difference, equal treatment stops making sense.

Reactivation looked successful for a long time

Long Read

We had multiple reactivation campaigns running in parallel.

Some were simple nudges. Users who had placed bets before but stopped logging in received boosts and reminders through email or SMS.

Others were operational. Dormant users were informed about wallet fees if they stayed inactive.

The most structured approach was AWOL campaigns. Users were divided into segments based on deposit behavior and reactivated with different boost levels.

From a distance, everything looked good.

Campaigns increased active users. The money often came back quickly. On a one-month view, results were positive.

That made it easy to conclude that reactivation was working.

And technically, it was.

But that conclusion was still too simple.