How accurate are SimilarWeb's revenue estimate numbers

[joemost]

I've noticed with SimilarWeb's update they now list an estimate revenue figure.

For example, this site here: https://www.similarweb.com/website/lineups.com/#overview

They estimate 1 Million Page views monthly, estimating 1 million dollars of revenue a year.

That seems kinda absurd. I was just noticing figures like this when looking around at my competitors. I'm guessing theyre calculations are based off what they thing ads would be worth on the site.

Curious if anyone has any insight as to how accurate these numbers could be.

 

[Philip J. Fry]

Plus or minus 2.5x IMO. Metrics like this vary a lot. There was a study where they analyzed the top 1,000,000 sites on the web list and found that the overlap was only 20% or so. For the top 1,000,000 domains! The Internet is huge. How a crawler (or toolbar or however SimilarWeb is getting its data) is programmed and what it indexes will give huge variations to the sample and results.

With that being said, I find SimilarWeb's traffic estimates to be better than Alexa's. You can't use the revenue estimates at all, IMO. Who knows what their margins are, CPMs are, click-through-rate are etc. That's a totally inaccurate guess.

 

[harrytwatter]

I work with some acquisition guys that use SimilarWebs traffic estimates and then extrapolate revenue estimates based off their own earnings. Obviously contingent on owning a monetized site in that space, but say you're in the fitness niche and doing competitor research on other fitness sites with SimilarWeb, you can break down your own EPM/EPC and apply it to the SimilarWeb traffic.

CRO is a hell of a differentiator though so in the end, impossible a high confidence level with these tools I think.

 

[ogmios]

Just took a couple of looks at some sites where I know the numbers for their revenue... Similar web is pretty much overinflated by so much that it is totally useless.
In comparison, AHREFs estimated traffic value is a lot closer and in line for the few I checked against.