I run a little tool for prediction markets (picount.com) that uses the twitter API to get the current “statuses_count” from three twitter profiles. During traffic spikes, this call will often fail and just return a blank value. When it happened this morning there were about 120 people on site hitting refresh, am I just getting rate limited? If so, what do I do to prevent this?
Is this what the Streaming API is for? Do I need to switch from a client-initiated query to a chron job that just puts these statuses_count numbers to a database that I can then read from?