Should NPS be reported by the date the survey went out, or the date the response came back? It doesn't seem an important choice, but it has implications for how the reporting and data can be interpreted in the business. Here are the pros and cons of each choice and why "received date" is usually the best answer.
Manual transactional NPS programmes survive about six months before someone gets bored of doing the data download and upload, and the programme quietly dies. Automation isn't optional. Automating NPS data collection requires: timely survey delivery, efficient reporting, in-depth analysis and closed-loop service recovery. We'll also cover how to choose between general survey tools, in-house builds, dedicated NPS vendors and CRM extensions.
A great score on a feedback attribute can be a great score or it can mean nothing more than that your brand is propping up a poor score. Psychologists have known about the halo effect for a century. Here's how to spot it in your data, with a worked example from a Canstar mobile phone survey.
Once NPS data is flowing, the question of targets follows almost immediately. Set them badly and you'll get score-begging, gaming and demotivated staff. Here's a practical approach to setting NPS targets: the prerequisites, local targets vs global targets, and why you should wait longer than feels comfortable before setting any at all.
CSAT has been around since the 1980s, so it must be the safe choice, right? Not quite. Customer satisfaction does predict loyalty, but on its own it explains only about 8% of it. What the largest study of satisfaction found, what a review of thirty years of research found, and which metric to choose when you can only run one survey.
Customer Effort Score was sold as the metric that would dethrone NPS. But new independent research from Groningen, with no axe to grind, runs the numbers and the result is more interesting than either camp saw coming.
Jump straight to the topics we write about most. Each link goes to that category's index.