Recently I was listening to a Redfin Earnings Call where CEO Glenn Kelman stated (somewhat vaguely) that homes sold using Redfin’s services "sell faster."

This stood out to me because just this weekend I was talking to a Real Estate broker in Orange County, CA (who competes with Redfin), who had (also vaguely) stated the exact opposite.

I decided to get to the bottom of things, and answer the question: “On average do homes sell faster or slower with Redfin?”

The verdict is surprising: It doesn’t matter.

Rather, this is a low value question, with much more noise than signal, which makes it the perfect battleground to squabble over, but not a good way to determine the quality of the product.

Disappointed with a non-answer? Let me explain.

Time on market is the perfect head-fake metric because it matters a lot to home sellers, and therefore is important for brokerages to wax loquacious. But it is much more important for the sellers to think that their broker is the fastest than that actually being the case.

This is because there is such a large distribution between how fast houses close, and the time on market is affected by a great many factors, but most have to do with the house itself and less have to do with the brokerage listing it.

Brokerages like to talk about how their advertising is the best, but when it comes down to it, the factors that matter more are the market the house is in, price, demand, seasonality, the neighborhood, the condition it’s in, and well, things about the house itself.

It’s kind of like asking the question, “Am I more likely to get a divorce if I have Sarah or Bob officiate my wedding?” You could probably find the objective data on divorce rates for Sarah vs. Bob, but at the end of the day the factors that lead to divorce have a lot more to do with you.

I reserve the right to change my opinion if I am presented with data that would suggest otherwise. But thus far, as I’ve researched the average time it take a house to close, most brokerages don’t prominently display specific average closing data.

For now, this seems to be an argument worthy of little attention.