Retention Is A Lifecycle Problem, Not A Bonus Problem
When retention numbers fall, the first instinct is usually to increase bonus spend. That instinct treats a lifecycle problem as a pricing problem, and it tends to buy a short recovery at a permanently higher cost, because players who return for an increased offer have been taught what returning is worth.
The lifecycle view starts earlier. The first session decides whether a player ever reaches a second one, and it is shaped by how quickly registration completes, whether the first deposit method works on the first attempt, and whether the product a player was promised in the advertisement is the product they land on. None of those is a marketing lever, and all of them are usually owned by different teams.
The middle of the lifecycle is where segmentation earns its keep. Players who arrived for a single product, players who move between products, and players who only appear around particular events behave differently enough that one message to all of them will fit none of them. Segments defined by observed behaviour hold up better than segments defined by value bands alone.
The end of the lifecycle deserves an explicit definition. Deciding what dormancy means, and what a reactivation attempt is allowed to look like and how often it may be repeated, prevents the reactivation programme from quietly becoming the largest and least examined line in the marketing budget.