CASE: testing whether a word was suppressing reach

CASE: testing whether a word was suppressing reach
An account suspected one term in their niche was throttling them. They posted 20 pairs — identical tweets, one with the suspect word, one with a synonym.
— With the word: median 1.9K impressions.
— Synonym swap: median 7.4K.
Label: Rumored (small sample, no confirmation the word is the cause).
What it means for your reach: you can isolate a suspected sensitive term cheaply with paired posts instead of guessing. Keep the sample over 15 pairs so you're not reading noise, and re-test before you permanently retire a word — these lists change quietly.