Beta is used for hedging (how much index to short against a stock book), for attribution (how much of your return was just market exposure), and for isolating alpha. All three uses depend on the estimate being stable, which it often is not.
Estimation choices change the answer substantially: daily versus weekly returns, one year versus five, raw versus shrunk toward 1.0. A stock can show a beta of 0.9 on two years of daily data and 1.3 on five years of weekly. Neither is wrong; they answer different questions.
Worked example: a $200,000 long book with an estimated beta of 1.15 carries $230,000 of index-equivalent exposure. Hedging it needs about $230,000 of short index, and if the true beta was 1.35 you are left with $40,000 of unintended long exposure.
Related: ols-regression, hedge-ratio, alpha, rolling-regression