28 March 2022

Customer feedback plus operational data equal a goldmine of insight

How do customers usually give feedback? How do we know that they are satisfied and they have had a great experience?

Usually, it’s with direct feedback - mostly surveys. You make a call to a customer service line and as soon as you are finished there is a question about how satisfied you are. These surveys capture customer feedback immediately after the customer has been served so they are fairly accurate - although the responses can be distorted by customers who are still angry about a negative situation.

Sentiment analysis of social media posts is another tool. I would argue that this is one of the most powerful because it allows us to monitor what customers are saying about products and services without asking or prompting. Obviously, in a survey, there is that question “what do you think about our service?” When monitoring social media you can see what people are posting without any prompts - it may be raw and unstructured, but it is real.

Many major brands have already dropped the formal survey approach and just use social “buzz monitoring” to track what people are saying about their brand and their products.

But I believe there is a deeper way to gauge what customers really think and that is a combination of surveys, buzz monitoring, and also operations data.

Think about it. You have data on what people are buying, what they are looking at, what they like, what they dislike, which products match with others. All this information sits inside your operational database. It’s the regular daily activity.

Why not modify your survey questions based on what you can see customers actually doing? Why not overlay your social media sentiment with the operational data to see if there are correlations? Can you see products that are not popular actively being discussed online?

These connections between the way that customers are actually shopping and behaving and what they are saying online is far more powerful than just asking a survey question. In fact, I would argue that this blend of operational data with primary research and sentiment analysis can be a goldmine for your business.

Surveys are often flawed. Customers might say what they think you want to hear, rather than what they really think. Sentiment analysis is closer to how customers really think, but it can also be skewed because most people will only comment online if their experience is really good, or really bad. We have all seen the social media posts about a terrible experience, and likewise, there are examples where people are so impressed by a brand they will write a blog or post something on their Facebook or Twitter.

All these feedback channels can be distorted to various extents, but the operational data is true. You cannot change the record of how actual customers have behaved when making real purchases. I think that we really need to start connecting this operational record with other forms of feedback if we want to create deeper insights into what our customers really like - and what drives them away.


Published first on LinkedIn by Jonas Berggren, March 2021.