How to Score Customer Service Interactions: Weighting a QA Scorecard

Once you have created your Quality Assurance (QA) scorecard - aligning your company goals with your scoring criteria - you're ready to build each section of said quality monitoring scorecard.

This article will cover the types of questions you can ask, as well as giving you an introduction to weighting your scorecards (as some questions might be more important than others). 

What Types of Questions Can I Ask? 

Different questions achieve different goals. Here are the 3 main types of answers used in quality assurance. 

Yes/No

This is a typical question you can ask in a quality monitoring scorecard. Expectations are clear, making it easy for agents and reviewers alike. 

💡Example: 'Did the agent do everything required of them in terms of content and procedures?'

This is ideal for when there are parts of your quality assurance process that are black or white - the agent either did it, or they did not. This specific question is a good one to ask when you pride your customer service on going the extra mile. 

Scale (e.g. 0-5)

There are blurred lines to rating how effective an agent's communication was so here we have a scale. In QA, embrace the grey areas, but do so in a data-driven way.

💡Example: 'Did the agent communicate with the customer in an accurate way? Did they display strong product knowledge?'

We could imagine that:

5: Excellent communication, no errors, did everything to solve the customer's query

3: Good communication, some grammatical errors but nothing major, did everything to solve the customer's query

0: Average communication, some major errors and/or did not do everything to solve the customer's query

To Note: 0, 3 and 5 is just an example - you can have each number between 1-5 if you want - but we always suggest that you make quality monitoring as simple as possible. The fewer possible answers you have, the more aligned you will be as a team - and the more consistent you'll be when scoring. This makes reviewing interactions easier, and makes your data more accurate, reliable and actionable. Calibration will ensure that these expectations are clear and that everyone interprets the scoring guidelines in the same way.

Written Answers

In this case, we have written wording for the answers. This helps make criteria clearer and perhaps less objective than the numerical values in the previous example (if there is uncertainty between what constitutes 3 or 5, for example).

💡Example: 'How is the agent's tone of voice?'

Again, it's a multiple choice answer - but if your team is more comfortable with written responses rather than numbers, this could be the way to go. 

The Key to Any Quality Assurance Program

No matter what answers to choose to your questions, you have to ensure that all of your reviewers are aligned. If you have a rating scale of 1, 3 and 5 for a question - they should be in agreement when it comes to what customer interaction would be a '1' when it is borderline '3'. 

For more on aligning your QA teams on customer service quality standards, download our Calibration Guide. 

The Importance of Weighting in Quality Assurance

It is certain that you'll consider some criteria to be more important than others. As we said, QA should be about checking criteria that lead to the improvement of a specific (business or departmental) goal. Let's have a look at several examples, taking into consideration the scoring criteria we named above.

💡Example: 'The most important thing is that the agent improves First Time Resolution rate

If this is your goal, the most important question in your scorecard could be: "Effectiveness: Did they minimize the likelihood of this customer contacting us again with further questions?"

Let's say there are 8 questions in your scorecard. You might say that having a 12.5% related to effectiveness is too little - so you can boost that section to be worth 30% to insist on its importance. In this particular case, perhaps the other 7 questions (like 'did they reply in a timely manner?) can be weighted at 10% each.

Using the example in this picture above, the top mark for a specific question (1 of the 23 questions being asked) is worth 10 points. The scorecard is out of 100. Therefore, we can simply say 1 question worth 10 points (10% of the score) and the other 22 are worth 90 points (90%). In this case, the company greeting is especially important.

💡Example: 'Our company tone is crucial to our brand - we must always be friendly and informal'

Communication, and the way your brand comes across, is your number one priority. You're a big believer in the power of word of mouth, and you measure NPS, you might believe the best way to interact with customers is to give them memorable experiences so that they tell their friends and family.

In this case, you want to give more weight to questions related to personalization, empathy, tone of voice and proactivity. Are agents going above and beyond for each customer?

In retail or e-commerce, teams may give more focus to the customer experience rather than an 'adherence to the process' section.

What Questions Should Be in a Quality Monitoring Scorecard? 

We've built a list of the most common topics/questions that we see in quality monitoring. Hopefully this will provide you with some good ideas as to which questions you should be asking in your scorecards to improve your customer service!

No two businesses are the same, so you do what feels right for you and the goals you want to achieve. Don't worry about cramming in every question under the sun - it's just not feasible! 

Want all of our Quality Assurance tips? Get the full Customer Service QA guide!