If you are looking for a way to find out how much traffic a website gets – the single best tool you can use is Similarweb.
As Google Analytics is the first step to see how your website is performing – SimilarWeb is the start point to find out how your competitors are doing.
Similarweb outperforms other competitive analysis tools in several aspects, including: accuracy of prediction, number of websites covered and amount of additional information. Of course to get a more complete picture it makes perfect sense to combine it with other tools – but if you have limited time, Similarweb is the best shot.
Below I explain in more details the strongest points of Similarweb, compare it to other web analytics websites, present its limitations and suggest how to use it with other tools to find out how much traffic a website gets.
What makes Similarweb outstanding from the rest in terms of traffic estimation:
1. More accurate predictions than other web tools – according to the research ran in early 2015 by Rand Fishkin (Moz’s co-founder).
2. Metrics that online marketers care about – visits, bounce rate, time on site rather than ambiguous ‘people’ or ‘global rank’.
3. Very strong data coverage – real data for huge number of sites, even the smaller ones (data for 87% of the websites in Fishkin’s sample were available, in comparison to poor 31% for Compete and dramatic 14% for Quantcast).
4. International range – offers traffic data from all around the globe with breakdown by country
5. Good in reflecting seasonal trends – high correlation between real and predicted traffic changes over time
6. Broad and useful additional information – including review of traffic sources, visitors breakdown by countries, engagement stats (time on site, pageviews, bounce rate), global / country / industry rank, basic mobile apps analytics
7. Relatively wide functionality available for free
Even though Similarweb currently seems to be the best competitive analysis tool available on the market, it is important to remember its limitations:
1. Accuracy is still far from perfect. In Fishkin’s research only in 22% of cases predictions laid no further than 30% from the actual numbers. It means that in majority of scenarios difference between estimate and real number can be bigger.
2. The smaller websites – the less reliable results. Accuracy of the prediction increases with the size of the website. For the smaller sites results are less reliable. It seems that Similarweb overestimates size of the smaller websites.
3. Only monthly traffic from the last six months is available in the free package.
4. Mobile Web is not included in Basic and Pro versions. It means that actual traffic of the website is likely to be up to few times higher.
Let’s compare Similarweb with other available tools and find out what it makes better alternative to Alexa, Compete.com or Quantcast:
1. Compete.com is limited only to the US traffic, which makes it useless for analysing websites operating internationally or operating locally anywhere else. It has much smaller database (in research run by Fishkin – Compete had real data only for 31% of the sample, while Similarweb for 87%) – in practice it means, that for many websites (especially smaller ones) you will end up with no information. Finally, the result you get from Compete.com is rarely precise (only 9% of predictions are no further than 30% from the real result).
2. Quantcast has even smaller database than Compete.com (only 14% of websites from Fishkin’s sample) and lower accuracy.
3. SEMRush takes into account only search engine traffic. However is very accurate for websites with vistors coming mostly from this source.
4. Alexa allows only size comparison between websites through Alexa rank, which in many cases is not enough to estimate the traffic of the website. Also it is less accurate in reflecting changes of traffic in time than Similarweb, Compete.com or Quantcast)
If you want to improve accuracy of your estimation, you may want to get some data from other sources. Below you can find few ideas for extending your research:
1. If the website is US focused, you can use Compete.com. It offers longer time period than free plan of Similarweb, similar sites and traffic sources.
2. If the website relies mostly on organic traffic, you can use SEMRush for assesing its size and analyze in more details its positions on different keywords.
3. Quantcast can offer interesting demographics data – comparison between the website’s audience and average for the Internet.
4. Search for brand name in Google Trends can give you idea of seasonality of the traffic with extremely flexible timeframes (ranging from hours to years).