Google scraping issues

Hi guys, I hope this is the relevant forum to post my question/issue:

I have been scraping google for years now. Using a lot of different proxy services. About a month ago I encountered for the first time an odd thing...

When querying Google from the server we often get captcha while using the same proxy setup in browser it shows clean results without any issues (no captchas).

Serverside we use nodejs and different user agents to do the queries.

It seems to us that Google can detect the traffic from the server and serve captcha while using the same proxy ip in browser (setup as proxy) doesn't trigger the captcha. Browser in incognito mode offcourse. I have tested it with Chrome, Firefox and Microsoft Edge, all giving clean results while the server query returns captcha.

Anyone an idea on how tackle it? It's obvious that only user agent isn't enough. Thanks in advance for any ideas!

From what Ive heard is that The loaded JS part is tracking your mouse movement. If it is not looking organic => whoops. Some time ago I wrote my own G scrapers... at this times, not even the frequency of requests was a problem but just the total amount per time. Seemingly this has changed now a little bit

One addition: I also implemented some function to solve the captcha when it appears. But when you are already on "yellow" the final strikes came always quickly ("we are sorry, but...").

I can see the JS mouse movement argument but I have tested in browser not using any mouse, results appeared normally.