Researching and improving customer satisfaction should be a contact process – it will ensure you really understand why customers like and dislike and you’ll be able to increase their satisfaction. You can’t expect similar results if you run a survey once and end with that. Even if it turns out customers are 100% satisfied and you can’t improve anything it can change anytime – customers’ habits and needs change, competition constantly works on offering as high quality of customer experience as possible, you might overlook some changes happening in your company etc. This is why you should constantly research customer satisfaction and look for possible improvements.

Before we begin, let me highlight one thing: your should aim to beat results from the previous survey, not to compare yourself against industry benchmarks. They can be misleading, especially when you operate in industries that traditionally report low customer satisfaction, like entertainment (more on customer satisfaction across industries: http://www.forbes.com/sites/ciocentral/2012/04/19/customer-satisfaction-by-the-numbers-an-industry-breakdown/#3e4341c6e30b). If you reach the average score of 77%, should you be happy? On the one hand, you’re similar to your competitors. But on the other hand, you’re still far away from providing similar quality of customer satisfaction to IT companies who score 95%. Why shouldn’t you aim to reach their level of delighting customers?

In this article, I will show you what are the steps of a successful customer satisfaction process.

Steps of customer satisfaction process

Define a goal and metrics

It’s easy to say that you want to research customer satisfaction. But you should be more specific before you start a survey. Define a core metric that you are going to track (for example, Net Promoter Score). You can also set success metrics related to the survey itself – for example, the number of answers you want to get or completion rates.

Hint: the core metric should be quantitative – either expressed in numbers (like NPS) or answers to a close-ended question. It will make tracking changes over time much easier and reliable. Open-ended questions are useful for digging deeper into customers’ opinions and collecting valuable insights.

Create a customer satisfaction survey

When you know what exactly you want to research, it’s time to create a survey. But before that, think what is the best method for running this survey. Should you use targeted website surveys, in-message surveys, questionnaires, or maybe call customers (but keep in mind that online sources are usually favored by customers)? After choosing the method (read this guide to learn how to measure customer satisfaction if you need some help), it’s time to create a survey.

The basic rule: keep it as simple and short as possible to maximize response rates. Also, surveys must be relevant – make sure customers are asked only questions they can know answers to. It usually means the need of segmenting customers. Example: ask different questions a new and repeat customer and use a different set of questions in a post-purchase survey you send immediately after purchase and in a survey you send half a year later.

Collect answers

People new to analyzing surveys often get excited after collecting a few answers. It’s understandable that but you should wait longer for results. It’s difficult to talk about a minimum sample size because it often depends on the number of customers you survey. For most of the companies, it’s impossible to reach statistical significance and a few hundreds of answers or even less must be enough. For a very small company, getting more than 100 answers can be challenging yet it shouldn’t discourage you from measuring customer satisfaction.

Analyze results

This is the core part of the customer satisfaction process – analyze the results to find out how satisfied your customers are. If you use only closed-ended questions, analyzing answers will be easy – you’ll see a breakdown of answers. If you used open-ended question you will need word cloud to quickly make sense of collected answers. While analyzing answers, think what they imply.

Turn conclusions into changes

You analyze results not to be smarter but to be able to make decisions whose aim is to improve customer satisfaction. Without this step, researching customer satisfaction or running any other kind of survey is mostly a waste of time. If customers complain about the quality of customer service, dive into this issue and think what exactly can be improved. Maybe all the people complaining contacted the same customer service agent? If people complain about the website – talk to the IT team how you can fix the issues. When changes are in place, it’s time to move to the last point of customer satisfaction process.

Survey again

When you implement changes inspired by answers to one survey, run another one to see if they worked. Customers might not like the solution or solving one problem can bring another one to the surface. It’s important to stop the old survey and run a new one, even if it is the exact copy – then you will be sure that answers are not mixed. Don’t be surprised if changes resonate differently with new and returning customers – often people don’t like changes.


The described process ensures you will research satisfaction of your customers in-depth and be able to increase it over time. Remember that your goal should be to provide as high level of customer satisfaction as possible, not to reach or beat industry benchmarks. Challenge yourself and analyze how improving customer satisfaction affects your bottom line.

Related resources: