Stuck in a Survey Rut? You need to be Conversational

September 13, 2017

Clarabridge Analytics
Clarabridge Engage
Contact Center
Customer Experience
Conversation Analytics

A few months ago, my family and I stopped for lunch during a long drive. The restaurant had a receipt-based survey that I decided to take at a later point in time. When I got around to taking that survey, I discovered that it was taking me too long to complete and wasn’t holding my interest. Bottom line, I didn’t finish the survey. I am 100% certain that we’ve all had a similar experience.

There’s a reasonable chance that the survey in question had been rigorously designed and tested, that all the branching logic was validated, and that a lot of effort went into gathering and cleansing all the data necessary to make sense of my potential responses. So, what happened?

The main thing that happened is social media. Social media and social customer service has dramatically changed our expectations for interactions with businesses, brands, and each other. It is immediate, it is free-form, and it has allowed companies to distinguish themselves in terms of responsiveness and accessibility. Many of us have become accustomed to tweeting at a company and getting a fast response, and this has reset expectations across the board. In the unstructured world of Twitter, Facebook, WhatsApp, and Google, pulling a customer through an ordered set of probing questions that requires focus and time has narrower applicability than it did in the past.

This reality is evident in the multi-decade decline in survey response rates, and has been exacerbated by the rise of Voice of the Customer programs that require hard data to measure progress and enforce accountability. If you’re getting fewer survey responses, then you’re getting less of the data that your various departments need for decision-making. As a result, the temptation is to ask more questions, so that those few responses contain more actionable data. Unfortunately, data quality tends to drop as surveys get longer (due to satisficing and outright abandonment).

The solution appears to be unpleasant at first glance: short surveys that deliberately ask fewer questions. The fear is that these shorter surveys will never be able to touch on all the KPIs that businesses find important. Caught between a need for more information and an inability to reliably collect that information via comprehensive surveys, companies have shortened surveys via methods such as branching. These techniques trade survey comprehensiveness for higher completion rates and more ‘white space’, focusing the survey down to several topics that the customer should have opinions on, but losing the ability to collect information outside of this funnel.

The solution to this problem requires a different approach. It turns out that you can have a shorter survey and still capture the data you need to effectively run your business. It requires thinking differently about KPIs, putting a premium on your customers’ unstructured thoughts, and applying the power of Natural Language Processing and Sentiment Analysis to the data you collect. When you do these things, your narrow and ordered set of questions becomes a Conversational Survey that truly gets at the heart of your customers’ needs, wants, and desires.

In this blog series we’ll cover the details and talk about how Conversational Surveys can help you effectively collect actionable data in a social media-heavy world. Talk to you soon!


Pete runs the Engineering group at Clarabridge and also manages the CX Survey product line.  He hopes your likelihood to recommend this post is greater than 8.