SimilarWeb is a Scam (..Here's What Actually Works)

Written by Dan
Read Comments
Share This Post!
similarweb.txt
Insight into how SimilarWeb works and why you cannot trust the data it provides.
When new advertisers and affiliate programs approach me about promoting their offers on my websites, I am often asked the same 2 questions: "How many visitors do you get?" and "What countries do your visitors come from?" Recently the routine has changed. The only question I get is "How are you? Then, instead of asking about my visitors, they start telling me about them!

Sometimes it isn't so bad. "We are looking for customers in the countries where you have visitors. We would like to make you an offer." More often than not, however, it is the dreaded "I can see you don't have lots of traffic..."

The latter claim simply just isn't true. This website and the others I promote receive many thousands of visits every day. Even if they did not, how would anyone know about it? I do not make the statistics of any of my websites public, and considering the value of this data, it would be foolish for me to do so.

What the advertisers and affiliate programs are doing is using a website called SimilarWeb to do "research" about my websites. It's very easy to use. They just need to type my domain name into a search box, click GO, and SimilarWeb will spew out some numbers and AJAX charts labeled as the "Global Traffic Rank", "Top Countries", and "Pages per Visit" of my website. The reports actually look very clean and official, but how accurate are they?

similar-bs.jpg
SimilarWeb publishes a lot of data, but our tests found it to be very inaccurate

The "Our Data & Methodology" PDF on SimilarWeb's site explains where their data comes from. Apparently it is a combination of 4 of their assets: a browser toolbar, a search bot, data from ISPs, and an algorithm to crunch the numbers. In theory, the toolbar alone could collect all of the data necessary, but theirs is not available for some browsers, and for those it is, less than 1% of Internet users even have the SimilarWeb toolbar installed.

To make up for this discrepancy, SimilarWeb uses their other 3 assets. The search bot crawls websites, analyzing them and grouping them into categories. The data from ISPs is not really explained, so it's difficult to know how much of it there is, or which ISPs are involved. As for the algorithm, it crunches the numbers that SimilarWeb have, filling in any holes in the data and multiplying it to produce their final estimated numbers that are displayed in the reports described above.

This idea is nothing new. A website called Alexa has been offering the same type of information since forever. SimilarWeb, on the other hand, is quite new, and that might be the biggest reason for its popularity with the folks I'm dealing with. Everyone already knows that Alexa's data should be taken with a grain of salt, and that it isn't particularly accurate. Also, its reports clearly state that the numbers are "Estimates" right where they are displayed. SimilarWeb has a fresh look that's kind of like Google Analytics. It seems to be trying to give folks an impression that its numbers are 100% accurate. The fact that they are actually just an estimate is hidden somewhere deep in the technical mumbo-jumbo.

So the big question is, how accurate is it? I decided to perform a couple of quick tests to get an idea.

Firstly, I compared SimilarWeb data and Alexa data side-by-side for 10 random websites. I logged the Global Rank, Bounce Rate, Visit Duration, Pages per Visit, and Top Country for each. The results were as I predicted: totally inconsistent. The data did not match up for any of the sites, and there were big discrepancies throughout. That doesn't prove much though. Perhaps SimilarWeb is accurate and Alexa is not, or vice-versa.

So, secondly, I compared the SimilarWeb data with my real statistics. When I focused on the Traffic stat, I found that SimilarWeb was reporting WAY fewer visits than my sites actually have. Its Global Ranking data was equally unimpressive, claiming that some of my most visited sites were less popular than ones that have no traffic at all. Its country-by-country data was particularly bad. In one case it claimed that Israel was accountable for 29.8% of my traffic, when in fact it only makes up 2.7% of it.

Why is SimilarWeb publishing this false information? Because they sell "upgrades" to their reports, which are likely stuffed with even more bogus information, starting at $200 per month. It's a scam, and not worth your time or money. Don't be fooled by the allure of SimilarWeb.

Related Articles

Top 5 Pros (and Cons) of Being an Affiliate
As a full-time affiliate I have experienced my share of ups and downs. Here are my pros and cons of being an affiliate.
affiliate-pros-and-cons
Are Celebrity Brand Ambassadors Actually Your Competition?
Affiliate programs sometimes ask you to promote their brand with a celebrity image, but it might not be a good idea to.
brand-ambassadors
Get a Real Job? Here's Why Successful Web Affiliates Are Great for the Local Economy
Affiliates can have a difficult time 'fitting in' where they live. I'll explain why our job is better for the local economy than most others.
local-affiliates
Revenue Rules of Thumb
Choosing between CPA and Revenue Share for the sites you decide to promote is a very important decision.
cparev
Pay Per Click Fraud Prevention
Affiliate guide to PPC Fraud and what you can do to avoid it
ppcfraud
Making a successful gambling affiliate site
Making a successful gambling affiliate site is no easy task. The people who have been able to make it work always put in a lot of effort. However, the online casino gaming industry is extremely successful. On that basis alone, people should be able to launch affiliate sites that achieve some degree of success, even if they are new to the field of affiliate marketing in general.
making-gambling-affiliate-site

Follow Us Now!

Comments on this Article

timesup

Premium CashMirrors User

"Worst part is how they say you can share your Analytics account if you don't like what the report says. It's almost like blackmail: either you share your stats with them or they put out a bad report about your site."

Posted: October 27th, 2016
Log in or Register to post your comments here.