Deliberately deindex a ranking page in three cases: when it cannibalizes a more important page, when it pulls in wrong-intent traffic that hurts conversion or sends bad engagement signals, or when its very existence drags down site quality despite the ranking. A page that ranks is not automatically worth keeping, because ranking is only one outcome and it can come bundled with costs that outweigh it. When the ranking does more harm than good, removing the page is the right call even though it is currently visible in search.
The first case is cannibalization. If a minor page ranks for a term you would rather have a stronger, more strategic page own, the two are splitting signals and clicks, and the weaker one is holding the better one down. Deindexing the lesser page lets the important one consolidate the ranking it should have. The principle to apply across all three cases is ranking-but-net-harmful: the question is not whether the page ranks, but whether keeping it ranked benefits the site on balance.
The second case is wrong-intent traffic. A page can rank for a query whose searchers are not who you want, and that traffic can hurt rather than help, tanking conversion, raising bounce, and feeding Google engagement signals that suggest the site disappoints visitors. The third case is quality drag: a thin, off-strategy, or low-value page that ranks for something marginal can still lower the site’s average quality and dilute crawl and focus. In both, the page’s ranking is real but the page is a net liability, which is exactly when removal beats keeping it.
The reflex to never touch a page that ranks is the trap, because it treats ranking as the only thing that matters. Run the net-harm test instead. For each ranking page you suspect, ask whether it steals from a better page, attracts the wrong searchers, or drags site quality, and weigh that against what its ranking actually delivers. Deindex it only when it does net harm, and leave the rest alone.