Support ticket classification is the grouping of your tagged support tickets to set you up to draw out unseen insights.

Coincidentally, doing it right is mission-critical for getting competitive value out of your customer support analysis approach. 

Your execution can make the difference between a great customer experience journey (aka customer experience  - CX) and a subpar one.

When implemented correctly, accurate insights - the downstream product of great ticket analysis - feed into your customers' needs and both drive up customer satisfaction (CSAT) and increase overall brand loyalty.

This rough-and-tumble guide will prep you to build your own accurate and efficient ticket classification approach in no time.

Here are the topics we’ll cover:

1. What Is Support Ticket Classification

2. Benefits of Support Ticket Classification

2. Why Support Ticket Classification Is Important 

3. Setting Up Your Support Ticket Classification Process

4. Automating Support Ticket Classification with Machine Learning

5. TL;DR - Takeaways

‍

Let’s get down to it 😤

What Is Support Ticket Classification

Support ticket classification is the process of adding tags to your customer tickets in order to reveal new, previously unseen data insights. 

But you’ve heard that (perhaps overly) technical definition already 🤓

Let’s take a look at a more approachable explanation.

Support ticket tagging is the second step in the customer feedback analysis process where you take tagged customer tickets (issues to fix, requests for help, etc)  and group them by new tags (think descriptions that characterize the text), revealing new information for your teams to act on for the purposes of improving customer experience (CX) going forward. 

A definition is one thing, but data can at times be both more illuminating and persuasive. Here are the numbers behind why it's important to get classification right 📈

Benefits of Support Ticket Classification

Here’s the cocktail napkin math for ticket classification success:

High-level ticket classification ➡️ effective ticket analysis 🟰 competitive, positive customer experience (CX) & customer satisfaction (CSAT)🫀

While his cocktail napkin math can seem simplistic, if you take each step seriously and execute, the results can be staggering.

Our CEO, Sharad, would be the first to point out how, even now, customer service analytics are often overlooked, even though they pose a massive value-add:

“Support conversations are a unique combination of easily available, highly valuable and yet... widely underutilised.

In many sizeable B2C & B2B brands, support tickets flood in continuously high volumes creating a large database of voice of customer data.

Whilst overwhelming for many, if harnessed quickly they become a real-time finger-on-the-pulse of the customer’s experience.

With a quality tagging taxonomy, the customer insight hidden away in the conversations opens the door for a wide range of use cases that prioritise experience and create efficiencies for support teams."

- Sharad Khandelwal, SentiSum CEO‍

Sharad's words illustrate an exploitable market inefficiency that can elevate your brand: Quality ticket analysis made possible by effective ticket classification.

Now, ticket analysis]is itself a subset of customer service analysis, which takes into account support tickets, reviews, social media feedback, and much more.

This in mind, while Sharad’s words might have piqued your curiosity, we’re ready to put forth an even bigger claim:

We firmly believe customer service analysis can be the most valuable source of customer insight 👀 🚀 💎

Ok, now let's locate where ticket classification ranks in all of this.

Ticket classification is an early upstream step in the ticket analysis process – and that's exactly the reason you should take great care to get your classification approach right.

🎯 Quality In -> Quality Out: Much like the mouth of a river, if you conduct effective, accurate classification, you ensure impactful end results.

‍
Amazing - let's say you're revved up and prepared to start upgrading your classification process – there's some quick housecleaning to do, then we're ready to get you started.

We'll hit the distinction between ticket classification and ticket tagging, then lay out some helpful resources for continued reading on ticket analysis at large, and, having checked those boxes, we'll be ready to crack on to building your new classification approach 💪

Ticket Tagging Vs Ticket Classification

Ticket tagging is the process of investigating customer tickets for characteristics and descriptors and adding those that are missing.

Ticket classification refers to the invention, application, and further sorting of customer tickets based on strategically chosen new tags.

If that seems squirrelly at all, don't worry - the terms are often used interchangeably.

The important thing to take away that they both relate to the same goal: Describing tickets by tags and using these tags to draw out insights.

If one was to drill down on the key difference, it would be that classification is typically used to refer to the creation of strategic new tags to build off of an existing tag structure.

Again we want to hammer home that, regardless of what stage of your ticket analysis journey you're at, accuracy throughout your implementation is be an absolute must.

To that point, here's a quick rundown of some helpful articles (similar to the one just linked) that we've compiled for you to keep on hand.

Ticket Analysis Resources

Before we get fully stuck involved with our classification practice recommendations, here are some detailed breakdowns we at Sentisum have concocted to support and inform throughout as you build your ticket analysis process 🧱

• Why: Your customer service tickets are your most important source of insights.
• How: Exactly AI ticket tagging can optimize and revolutionize your business
• The Numbers: Behind support ticket classification, from value add to macro best practices in podcast format and summarized.    

Browse those to your heart's desire, or strap in for the longer haul as we dive into our tips for setting up your new process, providing tips and outlining companies that realized success doing just this along the way. .

Brace yourself, we're about to introduce you to some new lingo - the 'tagging taxonomy' aka the product of your hard work classifying tickets.

Building Your Ticket Taxonomy: Ticket Classification Best Practices 

There's a new word - here's what it means and why we use it\.

What Is a Ticket Taxonomy?

A ticket taxonomy is a collection of categories into which customer tickets are qualified, or sorted, for the purposes of analysis, insight discovery, and improvement.

Each time you perform ticket ticket qualification you are creating a new ticket taxonomy.

This involves:

1. Taking stock of your current taxonomy: What tags are you already using? Are they working? What is missing?
   
2. Creating new qualifications. Most of the success stories below relate to new tags that brands thought up, that deepened their taxonomy and uncovered previously hidden insights.
3. Applying your tags. Our in-house tips below aim to advise you through this tricky, but critical step. Getting your tags right, and making sure your support teams can use them correctly determines how effective a taxonomy will be in its lifetime.
4. Checking your results: Check out our section on the merits of automated vs manual classification below, including a detailed case study of how we built an automated qualification system for a client.

During this next step what we’re up to more specifically is building new tags to deepen our data.

Generally, best practice is to think of what features are most important and/or what your previous results were lacking.

Kirsty Pinner, our very own Head of Content has a few ‘new tagging’ tips to live by

🔎Kirsty’s Tips #1: Think of tags that clarify the difference between issues:
‍
“Think about how your tags will be used in the future. I recommend thinking about tags as “problem tags”, “question or request tags”, or “feedback tags”.…doing this makes it really clear what the data means.

For example, if you create a tag specific to a problem, like “unable to checkout using Paypal”, then if you report a 25% increase in this tag’s usage this month you can be confident you know why.”

‍

Next, Kirsty offers a rule of thumb for those unsure if they have too many or too few tags – getting the right balance is crucial.

🔎Kirsty’s Tips #2: The tagging ‘Goldilock’s Zone’:

“The sweet spot for great insights and ease of tagging for agents is a taxonomy with 30-50 tags maximum covering the main problems, questions and feedback that arise."

‍

As a final offering, she leaves us with two more kernels of essential advice: thought’s on keeping your tags accessible and specific:

🔎Kirsty’s Tips #3: Keep your tags accessible and specific:

“Keep the names of the tags as close as possible to the customer language typically used so it’s intuitive for agents.

Avoid general tags like “What is order status”. They tend to become a catch-all for any kind of issue, so agents apply them quickly and move on.”

Many thanks to Kirsty for helping keep our process sharp. For further reading check out our aforementioned complete guide to ticket tagging, or our piece on tagging best practices. 

Or, if you’re in more of a rush (or just prefer examples)  take a gander at these four use cases from brands in diverse sectors - these success stories aim to illustrate the kind of difference-making new tags to aim for

1. 👚 Organic Basics: Step 6 Enrich your data - Adding ‘sales funnel depth’ tag - Case Study & Podcast

2. 🕺 Grindr: Step 4: Assess tagging quality - Reassessing legacy tags - Case Study & Podcast

3. 🇬🇧 U.K. Government Blog: Step 5: Create new tagging taxonomy - Crowdsourced classification - Support ticket analysis experiment

4. 🗣️ EveryoneSocial: Step 8: Maintain your taxonomy - Internal guide for agent training

With that done and dusted, it’s time to move on to making an essential choice. 

Automating Support Ticket Classification with Machine Learning

Should you upgrade your ticket tagging with machine learning, or keep banging away at it using the old, painstaking, inaccurate manual method?

(We definitely don’t think one option is far superior 😅)

Compared to a manual approach, automating your support ticket classification will result in massive, top-to-bottom improvement both process and results-wise

Feedback automation yields higher quality results due to improvement in three key areas: 

1. Accuracy
2. Scale
3. Response Time

To demonstrate this impact, we’ll look at a success story highlighting the improvement a global brand was able to make by drastically reducing - you guessed it - response time.

How James Villas Reduced Response Time 46% Using SentiSum Support Ticket Automation

Generally, while people are far better at dealing in the abstract and with each other, AI/machine learning is better at accurately and objectively handling large amounts of data.

For James Villas, a worldwide vacation villa rental business with properties in 60+ countries, this made automation an obvious must-add to its customer service (and feedback) processes.

"The challenge was to prioritize tickets so that urgent cases were handled as quickly as possible so that our customers could travel at ease."
‍
- Johannes Ganter, Head of CRM & Digital Transformation at James Villas

Villas needed to be able to handle their issues quickly and with consistency – so they turned to SentiSum to supercharge their customer service approach ticket analysis AI.

While a NLP-based machine learning customer service conversation approach made sense for handling things at scale (SentiSums AI can be programmed to easily handle 3,000+ properties and their related issues in real-time), the real secret sauce was in automation's ability to quickly triage urgent issues.

So, James Villas’s team signed up for a 30-day free trial with us, which included free consultations with our AI experts along the way,  to see what an automated build might look like. Here’s the workflow implementation they ended up developing with us:

Pretty darn spiffy, if we do say so – with a process that fit their needs and having built trust in our products, the Villas team was ready to fully integrate the automation 🤖

Luckily - and full disclosure this was a great selling point from the jump – SentiSum offers comprehensive integration with Zendesk, the best-in-class customer relationship management (CRM) platform. 

So the Villas team, who were already using Zendesk, just had to plug-and-play their new automated NLP build into their existing business software setup and - after some quick calculations - were ready to start treating their customer feedback with the full power of AI 🤯

The results, we’re proud to say, we’re staggering. Zendesk was able to keep track of and display, the effects of the upgrade in real-time 👀:

Over three months, average response time dropped 51%, beating out the initial dent of 46% – and that rate has continued to improve since. We mention this because at SentiSum we’re constantly interrogating our process in the hopes of constantly remaining on the bleeding edge - we wouldn’t suggest our customers do the same unless we truly believed in it.

In this case, SentiSum and James Villas were aligned in attitude when it relates to constantly seeking improvements –  Gartner, Villas’s Head of CRM explains that:

“The future roadmap of Sentisum looks incredibly exciting too. For instance, we can't wait to integrate translation features [these features are now live] and serve our sharply increasing international customer base.”
‍
- Johannes Ganter, Head of CRM & Digital Transformation at James     Villas

Hopefully this gets your wheels turning as to what automating support ticket classification can achieve. Regardless how sophisticated your current approach is we believe SentiSum (and automation in general) can simplify and re-equip any setup to better serve customers, making working with us not only enjoyable but a demonstrable value add.

‍

TL;DR - Takeaways

Building an effective, automated support ticket classification approach can be easy.

We know it might not seem so, but trust us – the face that you are here, thinking about support tickets in the first place means you are on the right track for the most significant value-add in the world of customer feedback.

Having identified this potentially massive opportunity for improvement, all that’s left to do is rebuild your ticket analysis approach from the ground up - keeping in mind the many details about classification we’ve checked off in this guide

If that sounds like a lot of work, we recommend reminding yourself of two things. One, the results will be worth it - you’ll see significant customer satisfaction, retention, and with any effort at ticket analysis improvement.

Second, there is automated ticket tagging software designed to help you available. Automation is the future of customer feedback because of its ability to handle data objectively and at scale and your incredible, personalized new feedback approach exists out there – and it wants to make your process pain-free and as impactful to your KPIs and ROI as possible 🙌 

To put a button on it (pun intended), great AI/machine learning-powered automation eliminates human error and ensures accurate, consistent, and deep insights - why not go start taking advantage of it today?

See what a top-of-the-line tagging and routing AI can do for you by scheduling a demo with a Sentisum Support Expert, or see for yourself by signing up for a free trial.

‍

‍

‍

Niall Ridgley is an AI and Sentiment Analysis expert. Really, he's writes about all things tech and tech-adjacent, looking to explain the now and explore the future. He hopes to bring more eyes (and hearts and minds) to SentiSum potential, and is happy to be with a brand that, in his personal estimation, is such a clear-and-obvious must add in it's space.

‍