Let’s say I need to retrieve mentions timeline 1000 Twitter accounts. Previously I would receive around 2000 mentions per day for those 1000 accounts.
I’m not paying $10.000 per month for the new Premium API to retrieve 2000 mentions per day. So I need to use the Standard REST API. Polling at rate limit would need me to generate 1 request every 12 seconds for each account (75 reqs/15 min).
86400/12 = 7200 GET requests per day per account, or 7.2 millions GET requests for 1000 accounts.
With at most 2000 requests returning a useful result (because I have 2000 mentions per day on average), that’s a “hit” rate <0.03% leaving me with more than 99.97% of those requests done only to waste CPU and bandwidth both for me and for Twitter backend.
I don’t see how it’s worth even for Twitter, is that how we’re supposed to do now?