TrendScout 50

We aggregate data for each site from more than ten analytical services and open databases: alexa, similarweb, google trends and others, enrich our data and process the information received using our algorithm.
As a result, we get a fairly accurate assessment of the dynamics of visits for each site, on the basis of which we make a rating.

1. First, we identify startup sites among the millions of sites in our database. To do this, we check its availability in various startup databases and weed out projects that are older than 2015.
2. Next, we determine the category of the site based on the information we collect about it: description, technologies used, site content and keywords.
3. The last step is to calculate the scores for each site according to our formula. The formula takes into account the following data:

• Number of site visits in the last month
• Number of site visits in the month prior to last
• Site growth in percent for 1 month
• Site growth in the number of visits for 1 month
• Site growth over previous months

The more points a startup site gets in our formula, the higher its place in the ranking. Thanks to its formula, the quality of projects in the top improves.

Now the data completeness is at the level of 75-80%, i.e. we may miss out on some growing sites or startups, or categorize them incorrectly. But every month the algorithm becomes more accurate and loses sight of fewer projects.

The presence of the site in the top definitely indicates that the site has received a strong increase in visits. In about 10% of cases, there may be an error of more than 30% in the exact estimate of the increase in visits (i.e. we can estimate an increase of 500%, but in fact it was 750%).