The Big Survey Trade-Off

Historically, firms conducted market research by designing highly structured surveys, making assumptions around what is important to their customers and ultimately their business.

Fizzback typifies the art of the possible in contemporary customer engagement. Freed from the confines of older market research methodology with its traditional structure and low response rates, the ability to ask for open-ended verbatim feedback and have that data automatically computed into actionable sentiment and topic insight can revolutionise the way that companies engage with their customers. Response rates and the quality of the feedback are improved further as firms begin to solicit the feedback directly from customers at the ‘moments of truth’ in their customer journey.

Today, the big question our clients face when designing surveys to capture customer feedback is the trade-off between volumes & granularity. This comprises whether a firm wants to assess a suite of KPIs by using specific numerical questions or whether the firm leans towards open-ended response. The granularity of numerical responses and ability to track pre-defined and business-specific KPIs provides the main advantage of the former whilst categorized open-ended feedback and a valuable window into front-line customer insight is the appeal of the latter.


Fizzback undertook cross-client research across almost 1 million customer responses in 2010 in 5 countries and 3 industries to provide some thought leadership into the ‘survey trade-off’. The right end of the scale where clients opt to ask a single numerical question and a verbatim request enjoy a response rate of 69% on average. Compare that with firms asking 5 numerical questions in addition to a verbatim and the response rate plummets to 35%. Moving through the scale we found that, on average, each supplementary numerical question comes at the cost of an 18% drop in response rate.

So we see that the common temptation for firms to introduce extra numerical questions to their surveys to add more granularity and customer insight into their results may in fact be counter-productive. With the sagacious results we can derive from comments processed through Fizzback’s Natural Language Processing engine, it may well be the case that less is more when it comes to survey design.

Share this:
Twitter LinkedIn Facebook Email