Using machine learning models we automatically remove gibberish and irrelevant answers to your open text questions, so they won't be included in your final sample and results.

We layer a final manual check of the results on top of these automatic measures. If we remove and remaining poor quality answers after the survey closed, it will go live again to replace those removed with final quality responses.

We are also constantly implementing ever more sophisticated ways of eliminating bad participants, to reduce the number of poor quality answers in the first place. 

If you have any more questions please don’t hesitate to get in touch with our Client Experience team.