How do the services work: megaindex, spywords, serpstat?

Question

contextual advertising

Good day.
Please tell me how the services for collecting keywords for advertisements work. When I enter a domain name in megaindex or serpstat, I get keyword statistics (ad text, approximate budget, query, impression statistics). How do such sites collect this data? Judging by the response time, I assume that they already have parsed queries and they are given directly from the database. Or is that not how it works?
I did not find information on the Internet on this topic. I would really appreciate a constructive response. Thank you!

Answer 1

1. Analytics is collected in advance by the parser from public sources: counters (all the necessary indicators for the forecast) and search engines (number of domain mentions, topic demand, etc.).
2. Let's assume that the percentage of domains with open data is in the confidence interval (out of all that were parsed). When forecasting, we take into account only them, the rest are not touched yet.
3. Using CTR collected over the entire period of history, the predicted cost per click on any domain is calculated, taking into account the following parameters (everything is collected from the history of similar sites): number of impressions, number of visitors, business area / direction / topic of the site, average number of visitors per day/per month, percentage of competition (for advertisers), etc.
4. Calculate the same indicators based on the dependencies obtained for a smaller proportion of domains (they are in the non-confidence interval) and also write them to the database.
5. All this is constantly changing automatically in their database based on constantly changing statistics (time passes, domains change owners, HYIPs appear and disappear, etc.).
6. When requested, they already have everything calculated in the database and they immediately issue a ready-made price. It is important to note that the price can be now - one, and after 5 minutes - already another!
Something like this...

Answer 2

Yes, data is collected in advance, all requests are parsed in a row, statistics are collected on them