Top 5 Intercom conversation tagging challenges
Intercom is a powerful tool for business growth, allowing businesses to communicate directly with their customers. It means you can quickly and easily offer customer support, engage with users, and use it for lead generation
Intercom’s Business Messenger enables companies to communicate directly with customers, allowing them to track and solve questions and issues. It can build a valuable bank of data to inform business decisions moving forward, based on information that comes directly from users.
Various teams need insights into customer issues and needs, from the C-suite to product managers, as well as designers, researchers and marketers. With this data, stakeholders across the business can make better decisions based on customer feedback.
To keep track of consumer data, most customer support teams using Intercom conversation tags. This means they can categorise customer issues and requests to pass onto other teams across the business. However, most support teams are tagging issues manually, which leads to a plethora of problems.
Here at Prodsight, we’ve spoken to hundreds of customer support teams over the last two years about the issues they face when it comes to recording and maintaining data around customer requests. The same issues keep coming up again and again. We’ve summarised them below – do you recognise any of these issues from your own experiences with Intercom tagging?
1. Inconsistent Intercom tagging
One of the most common issues raised by the hundreds of customer support teams we’ve spoken to over the last couple of years is that tagging is inconsistent across their team.
Usually, support teams will have a clear framework for tagging support conversations. However, there are inevitably times when a customer support agent will tag a conversation differently from how you might expect.
That’s because conversation tagging is subjective. Even with a clear set of rules, agents might not tag a conversation according to conventions, because they perceive the issue differently.
Take, for example, a customer who initially reaches out with a question about their subscription, but then asks for support with payment. How would you tag that conversation? With a manual tagging system, two customer agents might have a different perception of how to track that conversation.
We’re all prone to biases when making judgements and decisions, and so it’s impossible to be sure that all conversations with a particular tag are actually related to the same concept.
It might not seem like a big issue to have a few conversations which are tagged differently, but it can really impact the accuracy of your reporting. It’s particularly important as your business grows, as large-scale inaccuracies in your Intercom analytics could lead to making business decisions which are based on wrong data.
2. Poor tag coverage
Tagging conversations inconsistently is one issue – but not tagging them at all is an entirely different one, with consequences that are potentially more far-reaching.
Unfortunately, this is one of the major issues customer success teams have reported to us when it comes to Intercom tagging – that is just doesn’t happen at all.
There are numerous reasons for this lack of tagging:
- Naturally, customer service agents are focused on helping their customers. They want to support them as quickly and efficiently as possible, which means they don’t want to take up precious time deciding how to tag a particular conversation when they could be helping the next customer.
- Marketing and product teams may have different perceptions of what’s an important insight. It might be difficult for customer support agents to visualise what sort of data these teams want to see from support conversations, so they don’t attempt to tag them.
- In instances where it’s unclear what tag should be applied to a particular conversation, service agents often leave them untagged. Although they will do their best to apply tags to support conversations based on the agreed structure, it’s easy to overlook conversations which don’t neatly fall into a clear-cut category.
All of this has an impact on your Intercom analytics and reporting. Some issues will be underreported, and others will be overreported, leaving you with inaccurate data that makes it difficult to glean any insights from.
3. Cost of manual Intercom tagging
Surely there’s not a significant cost associated with adding simple conversation tags to customer support conversations? You might think so, but actually, that cost can really add up when you consider that dozens of customer support agents are applying tags to thousands of conversations on a daily or weekly basis.
The cost will depend on the size of your business and how many staff are in your customer support team. But let’s imagine that you have five agents in your team, who spend half an hour every day tagging conversations.
The average Customer Support Agent is paid $17 an hour. So over the course of a year, five agents spending half an hour every day on conversation tagging would cost your company $10k. Not such an insignificant cost after all!
And that amount doesn’t even take into account the cost incurred by agents spending time on tagging rather than helping customers more quickly. When you take that opportunity cost into account, the amount of money saved – and the amount of extra revenue your business could earn – through a more efficient Intercom tagging process is really something to think about.
4. Inability to change tags retrospectively
This is a common issue when it comes to support conversation tagging. The purpose of tagging conversations is to gain insights into the customer experience, and understand what they’re struggling with in order to improve in the future.
However, it’s only upon researching and prioritising customer problems that businesses really learn what the overarching issues are. After digging into the problems flagged by customers, you might want to test theories in order to do some more investigation around particular issues.
For example, you might want to look into how many times a customer has mentioned a particular issue – but you can’t, because the tags have already been applied at the point of the conversation, and it can’t be changed. This is frustrating and doesn’t let you test and learn from the data to apply the insights when making any sort of decision to improve the customer experience.
5. Limited reporting
What’s the point in tagging conversations? Support conversations are tagged in order to learn from customer complaints and questions. Data around the consumer experience is truly valuable for businesses when making decisions and improving the customer experience. There’s no point in tagging conversations if they can’t be used to produce insightful reports to drive actions across the organisation.
Intercom is really useful for customer engagement and support, and whilst it has some great basic reports on conversation tags, it was never designed to be an advanced analytics tool.
That means that businesses can use Intercom to gather top-level insights on their customers’ experiences, but it’s difficult to uncover deeper insights. For instance, you might want to find out how many of your users on a pro plan, who signed up in the last three months, have mentioned cancelling their account during a support conversation. But that’s not something you can do with Intercom’s built-in reporting facilities.
It’s possible to do some clever things with Excel or Tableau to get all the answers you want from your Intercom conversation tagging – but who has the time to do that? Creating a manual data analysis process alongside manual tagging is time-consuming, and another challenge for your teams to stay on top of.
Try Prodsight to fix your manual Intercom tagging issues
A manual tagging process for tracking customer support conversations on Intercom is far better than no tagging at all. However, manual tagging does come with several issues, as reported by hundreds of customer support teams and outlined above. These are:
- Inconsistent tagging
- Poor tag coverage
- Cost of manual tagging
- Inability to change tags retrospectively
- Limited reporting abilities
To address these issues, we created Prodsight. We automate manual conversation tagging so that your team can spend their time helping your customers and creating informed solutions to their problem, rather than worrying about how to categorise a particular conversation.
Prodsight automatically discovers topics in your conversations, and measures sentiment, so you can quickly and easily discover what your customers need support with, and how they feel about your company. We also collate your mentions in one place, to make it easier for you to delve into the data.
It’s an automated system, meaning it’s very consistent. You can get statistically significant insights from the data we collate, without the manual labour, meaning your team’s time is freed up and you can make informed business decisions without any worries about data inconsistencies or inaccuracies.
Thinking that it’s time to make the switch from manual tagging to a more sophisticated system? We offer a free seven day trial of Prodsight, so you can find out what we have to offer and how it can improve your business, without any risk.
Just register for your free trial today, and get access to all of our features for a week, including topic and sentiment analysis of your customer support conversations.
Tadas is the Founder and CEO of Prodsight. He’s obsessed with the customer experience and data-driven product management. On weekends you’ll find him drumming, cycling, or drinking coffee.