Generally speaking, the more words the better.

Relative Insight can analyze text data sets as small as 200 words but for practical purposes, we recommend a minimum of 1,000 words in order to generate meaningful insights.

As an example, when analyzing social listening data we generally recommend 1,000 tweets per data set as a good-sized sample. This would amount to approximately 15,000 words. 

2,000,000 words is the recommended upper limit. Beyond this, additional volumes of text are unlikely to bring about new insights. 

Keep in mind that larger data sets may take more time for the platform to process and analyze.