What can explain analtytics search counts?

My search operations count seems really high. I can’t explain difference between global daily search operation count (46k)

Screenshot from 2020-06-08 12-02-50

and the sum of all counts I get from all different indexes:

Screenshot_2020-06-08 Algolia Dashboard(2)

There’s a huge difference between the two numbers (more than 50%). I think it cannot come from indexes updates as they are counted as “record operations”:

Screenshot from 2020-06-08 12-09-40

What can explain this difference ? Thanx for the help !

Hey,

thanks for your question!

Our Usage basically reports the number of calls that the Search API received.

As we strongly encourage building “search as you type” experiences, this would lead to a lot of noise if our Analytics reported raw queries. As a consequences, our Analytics reports “Searches” which result from a keystrokes aggregation (see this short documentation).

If the end-users are refining their queries a lot or searching for long queries, this can explain the difference.

I hope that helps!
Please let us know if you have further questions or feedback.

Marc

Thank you, it makes perfect sense ! I’ll try to reduce the amount of queries with debounce or something and see how does it look in the analytics.

Suggestion: you could show these numbers of raw queries for each index somewhere, it could help !

Thank you for the suggestion, I’m logging this!

We actually display per-index Usage metrics in the indices > stats section which is not always obvious. Does this suit your needs?