Customer feedback

How to manually analyze customer support tickets

Ben Goodey
Ben Goodey
November 30, 2022
Join a community of 2139+ customer-focused professionals and receive bi-weekly articles, podcasts, webinars, and more!

This is based on a chapter from a full ebook on the topic of driving growth from the contact centre. The original research is based on interviews with eleven industry experts. Download the full support-driven growth ebook for free here.


The customer services department is an asset. No doubt.

You make or break a customer’s experience with your brand and are the frontline during critical moments for the customer.

Support tickets flood in every day. Customers want help and often need it urgently.

That’s why every support ticket, whether it’s a live chat, email or phone call, holds an unparalleled insight into what’s causing agg for your customer.

“Agg”, otherwise known as reasons for complaint or customer pain points, is vital to share back to the rest of your business so that they can improve, remove friction points, and, ultimately, tame the causes of customer churn.

Support tickets are the purest form of customer insight. Unlike surveys, the customer comes to you. They want their problem solved and are more candid than ever in the pursuit of a solution.

Mining those tickets for customer insights can be critical for businesses who want to remain in tune with their customers and keep them spending more, for longer.

But analysing them can be an overwhelming task. High volumes are powerful when it comes to uncovering statistically significant, actionable customer insights, but when trying to make sense of them manually, it’s an often insurmountable task.

We’ve written a lengthy guide to customer feedback analysis, and now we’ve got the opportunity to go quite a lot deeper into support ticket analysis.

As a company that does this every day, we’ve built a strong knowledge around it. We automate the task using artificial intelligence, but that doesn’t make sense for everyone.

This guide is about manual customer support ticket analysis. We recommend it for those of you with 200 or fewer daily tickets—any more than that is a heavy daily task for anyone.

Export your support tickets for today to have a go at this. Then you’ve covered the ‘what’ to collect and the ‘how’.

Now we’re going to dive into how to turn your support ticket conversations into transformational customer insight. You’ll walk away with a small guide book with which you can demand change from the product, ops and marketing teams.

Let’s go.

AI for customer service ebook

4 steps to manual customer support ticket analysis

First thing that's important to know?

If you have thousands of tickets, this is going to take you AGES.

It's much simpler (and affordable) to automate this entire process with our AI-based conversation tagging and insights technology.

For those who endeavour to go on, let's get stuck in.

Collate your feedback data—Free template

Create a spreadsheet (click below to instantly download a template) with the support ticket logs alongside that customer’s key metadata (like average order value, length of time as a customer, date of feedback and source).

Download the free template here.

Your support ticket management software should help you uncover the metadata, and you can export your support chat logs as qualitative feedback.

Decide on your taxonomy

Categorising feedback isn’t always easy. Choose topics that are granular enough to be insightful but high level enough that you can draw out similarities. When dealing with high volumes of unstructured free text (such as customer support ticket logs) you’ll want to create both high-level themes and more granular sub-categories. 

Start with feedback type so that you can deliver the feedback to the right team. Here are some examples:

  • Feature request 
  • Product bug 
  • Third-party provider issue

Now go further to provide granularity. Here are some theme examples with sub-categories beneath them (this is called multi-level ticket tagging):

Delivery Packaging:
  • Delay
  • Damaged Packaging
  • Tracking
Order process:
  • Discount not working
  • Failed payment
  • Credit not applied
Ease of payment:
  • Payment provider failed
  • Card expired
Website issues:
  • Product availability
  • Display issues
  • Slow load time
Staff service:
  • Rude staff
  • Helpful customer service

We suggest scanning through a handful of customer conversations to get a feel for your themes and categories. High volumes are important for actionability, so you’ll want to research beforehand to ensure you cover all bases.

You should also keep cross-functional teams in mind. Ask them, what do they need tracked? What issues are they worried about and would like counted?

Start tagging the data 

Now that your data is in an excel spreadsheet, it’s time to get stuck in. We suggest one person doing this task to ensure objectivity—two people may interpret the same piece of text differently, categorising it differently.

Analyse row by row, and add the category and theme for that data point in the spreadsheet. Use the tick box method so that each conversation can be tagged with multiple themes and topics. Being a rich source of customer feedback, your customers are likely to mention multiple causes of customer friction in each ticket.

This task can be time-consuming with high volumes of data. It’s especially time-consuming if you choose to expand your research beyond customer support tickets, to survey analysis results and customer reviews. Because actionability is a key consideration, you are likely to only be able to analyse small ticket volumes in a way that’s timely and accurate.

For large volumes of tickets, again, we suggest using support ticket analytics and ticket routing software like SentiSum.

Look for patterns and share your insights 

Apply filters to your spreadsheet to reveal which topics are mentioned in the highest volumes, and send the data off to the relevant team. The popularity of the theme and category can be a clear indicator of what needs prioritisation.

When it comes to sharing your findings, who, how and when must be considered. Insights reporting should be simple for anyone to understand, so make sure to label the meaning of your chosen taxonomy. Share your findings as quickly as possible to ensure the data is more actionable, and add the volume of topic data to prove statistical significance.

How do we share insight?

Sharing insight is vital to it getting used. You're likely past the stage where you can turn around and tap the head of product on the shoulder and say, "dude, this needs to change!"

At SentiSum, we've created an email template you can use that ensures the right people are informed on a daily basis. Often that's all you need to open a conversation to what needs to change.

You can create a similar template using a simple software like MailChimp or Hubspot. We'll breakdown what goes into ours so you can consider what suits your company best.

Here's a screenshot:

SentiSum Daily Digest Customer Insight

Because it's an email, and a daily one, we try to keep it scannable. Limit the email to the important changes and let the reader dive deeper when their interest is piqued.

Here's what we include:

  • Top drivers of customer contact yesterday. Often these won't change looking from a longer term perspective, but day by day can show useful anomalies.
  • How each of those drivers has changed compared to their usual baseline. Benchmarks are important has a reference point, so you may want to kick off your analytics project by looking at your historical support ticket logs.
  • Ticket volume on each topic: a % change with volume lacks important context.
  • Biggest changes: With this section we have an opportunity to get more granular on what's driving customer friction. Sometimes a new topic comes out of nowhere, or a relatively infrequent topic experience a large swing in the wrong direction. This is important to keep track of because it's likely indicative of a new, potentially solvable problem facing customers.

Other additions to consider:

  • Note the channels with the highest volumes
  • Add notes to give context to the reader
  • Tailor the email to the team it's sent to (every team wants different information)

Our clients add lots of users to our daily emails. There's nothing better for encouraging customer-centricity than a daily email with customer insight in it. Especially if the team can see the daily impact of their work and continuous improvement projects.