When to retire an offer instead of discounting it again

An offer that used to convert stops converting. The reflex is to discount it. That works for a while, then it stops working, so the discount gets deeper. Then the offer gets bundled with something else. Then it gets a timer.

Each of those moves buys a smaller lift than the one before, and by the fourth round the offer is doing something worse than not selling: it is training your players.

What repeated discounting teaches

Players notice patterns in a store faster than studios expect them to, particularly the engaged players who visit often — which is to say, the ones most likely to buy.

An offer that appears at full price, then at 20% off, then at 40% off teaches a simple lesson: waiting is rewarded. A player who learns that does not stop buying. They stop buying at full price, which is worse, because it applies to the rest of your catalog too. The behaviour generalises. Once someone has been rewarded for waiting on one item, they wait on the next one.

The second thing it teaches is that the original price was not real. A price that falls repeatedly without the contents changing tells the player the number was arbitrary. That undermines the anchoring the rest of your ladder depends on, because anchoring only works when players believe the anchor.

Neither effect shows up in the discount's own performance report. The discounted offer converts better than the undiscounted one did — that comparison always flatters the discount. The damage lands somewhere else, later, and gets attributed to something else.

Why the offer stopped working

Before deciding what to do, it is worth being specific about why conversion fell, because the three common causes have different fixes and only one of them is price.

The first is saturation. Everyone who wanted this thing has bought it. The offer is not underperforming — it has finished. Its addressable audience is used up, and a discount reaches the people who did not want it at any price, which is why the lift is small and does not last.

The second is that the offer has been overtaken. Content shipped since it launched has made it less useful, or a better item now exists at a similar price. The offer is worse than it was, and the honest response is to change what is in it rather than what it costs.

The third is fatigue in the presentation rather than the product. Players have seen this offer so many times they have stopped registering it. Conversion falls because attention falls. Repricing something nobody is looking at does very little.

Discounting only addresses a fourth cause — that the price was genuinely too high for the value — and that is the least common of the four in a catalog that has been running for a while. If price were the problem, conversion would have been weak from the start rather than declining over time.

What retiring actually does for you

Taking an offer out of the store is not a loss of the revenue it was still producing. That revenue is usually small by the time this conversation happens, and some of it moves to whatever fills the slot.

The gains are less obvious. Store space is finite in a way studios underweight — a player scans a limited number of things before deciding, and every slot occupied by something they have already declined reduces the chance they see something they would accept. A catalog that only accumulates gets harder to shop, and the offers most likely to be crowded out are the newest ones, which are the ones you most need to test.

Retirement also restores a tool. An offer that has been absent for a season can return at full price and read as new, because the players who saw it worn out have mostly moved on and the ones who remain have not seen it recently. An offer that was discounted into the ground cannot come back at full price without contradicting itself.

The third gain is diagnostic. If you retire an offer and the revenue it was producing simply reappears elsewhere in the catalog, the offer was not creating demand — it was capturing demand that existed anyway. That is genuinely useful to know, and you cannot learn it while the offer is still there.

See how we manage offer lifecycles →

A workable rule for the decision

The question worth asking when conversion falls is whether the offer is still reaching people who have not seen it.

If most of the players seeing this offer have seen it many times before and declined, the offer is finished for that population. Discounting it is an attempt to convert people who have already answered. Retire it, and bring it back later for players who have joined since.

If the offer is still reaching new players and converting them, but the aggregate number is falling because your player base has more veterans in it than it used to, the offer is fine. The metric is misleading — it is averaging a working offer for newcomers with a used-up offer for everyone else. Segment before you act.

If the offer has been overtaken by newer content, change its contents. Same price, better items, and treat it as a new offer rather than a rescue of an old one.

Reserve discounting for the case where you have a specific reason to believe the price was wrong — usually because a comparable offer at a different price converts noticeably better among the same players. That is a price problem, and a price change fixes it. Everything else is a different problem wearing a price problem's clothes.

The general principle is that a catalog needs a way for things to leave it. Without one, every offer that ever worked stays in the store forever at a declining price, and the store slowly becomes a record of things that used to convert rather than a set of things a player might want today.

Talk to us about offer testing and retirement →