Chris Webb deals with management:
As you’ve probably seen in the blog post for the September 2026 release of Power BI, Import mode and Direct Lake mode semantic models now support the ApproximateDistinctCount() DAX function. I tested its performance on a Direct Lake semantic model with a 1.4 billion row fact table based on the NYC Taxi sample data and as you would expect, it was a lot faster than doing a regular distinct count in most cases.
Chris notes that the primary challenge is human: getting people to understand that, for their purposes and at the scale in which approximate distinct counts makes sense, non-biased approximations are just as good as actuals. That’s because 119,374,399 and 118,998,706 generally aren’t distinguishable in practice for things like monthly unique user counts. In both cases, the actual business user would still write “120 million” (or maybe “119 million” to be more precise).
Now, in cases in which the precise number does matter? Absolutely use the real distinct count. But those scenarios are rarer than you’d think.