Yes ablation as in removal. They then refer to using the 3% experiment to estimate out revenue loss, so it appears they were running an AB test where it was removed for 3% of users so they could understand the metric impact of removal. Part of being a decent PM is monitoring SM for any feature talk.