Learning From Users

Conversation mining involves mining or digging into conversations that your users are having about your product.^[1] These conversations usually happen in the form of reviews, social media posts (such as tweets), group posts, or comments, all within the digital arena.

These conversations can give you valuable insights about:

• How your users feel about your product
• What they like about your product
• What they dislike about your product
• What they’d like to change about your product

Keeping that in mind, conversation mining can be a great tool to optimize your UX copy and effectively attain your UX writing goals.

Conversation mining can be done on any digital domain where information is shared freely by users such as e-commerce websites, the review page of your website, social media sites, forums, and groups where users are active and vocal through their reviews, posts, and comments.

If your team has already designed a user persona for your product, this step should be fairly easy. Even if you do not have user personas at hand, use the data available to you to figure out where your users are present online. This will function as the source for your conversation mining.

Keep in mind that it is unethical and also possibly illegal to intrude into private conversations without users’ knowledge to carry out research for your product development.

Let us assume for a minute that your product is an online shoe store, that your conversation mining source is X and you’re mining to understand users’ current sentiments on your latest launch of winter boots.

Sure, people are talking about your launch online. But what types of conversations are these? Conversations about your latest shoe launch could be positive, meaning that you’ve done a great job. Or, it could be negative, showing you that there is still room for improvement. Or, it could be an equal amount of positive and negative, meaning that your audience is diverse and has varying preferences.

Conversations can be addressed directly to your brand or conversations about your product can take place without you. A user could simply comment about your launch, another could possibly tag you to get you involved in the conversation. Each type of conversation is uniquely helpful and can provide you with valuable insight into your product.

Approaching UX writing by mining positive conversations is a good option if you are writing for your product from scratch. It can be beneficial to understand what has worked in the past to set it as a benchmark for the present and future. Besides, it’s always nice to know what users really like about your product in order to really capture its essence. Positive conversations answer the question of how users view your product and what part of it speaks out to them.

In an attempt to understand what users liked about their $150 segment speakers, Amazon carried out a conversation mining study on its product review pages. Through this research, they were able to pin down the exact features of these speakers that appealed the most to users, and as a result, they could accurately pinpoint the factors that directly led to purchasing decisions.^[2]

Similarly, finding out what users like about your product can help you project it in a way that is most likely to drive them to take the actions that you want.

Neutral conversations are tricky to handle — they aren’t really positive or negative. Usually, such conversations are devoid of polarizing language, emotion, and opinion. This is mostly the case for information that is being reported, as in the case of news.

However, it is possible to draw insights from even seemingly neutral conversations. Here, you would not focus on whether the conversation is positive or negative, but instead on what the conversation is about. What about your product made it to these conversations? This will give you an idea of which parts of your product are prominent enough to attract attention.

If positive conversation mining helps improve your product content, negative conversation mining helps protect your brand image and reputation. Learning what your users dislike or detest about your product can go a long way in ensuring that you take corrective steps and avoid repeating these mistakes in the future.

In what was one of the greatest UX writing fiascos of the year, The Seattle Times reached out to readers with a notification asking them to subscribe for COVID-19-related updates. However, the CTA on their subscribe button said “Yes, I want Coronavirus,” leaving users perplexed. Many of these users took to the internet to express their amusement and anger. Through conversation mining, the issue was quickly resolved by another UX writer.^[3]

The lesson to take home here is that negative conversations need not be looked at as a bad thing because they have the power to pinpoint what is wrong. They can help you arrive at solutions accordingly, while also allowing you to stay away from such errors while writing.

Your users are like your North Star — let yourself be guided by them. This is especially true in the case of conversation mining, which can give you exceptional insight into the way your users think and talk about your product.

More particularly, it gives you a peek at their vocabularies and the exact words they use to describe your product. As a UX writer, this can help when you attempt to mimic your users’ language and thinking. For example, let's say your app is a fitness tracker, and after conversation mining, you notice users commonly refer to their progress as “crushing goals” or “smashing workouts.” Instead of using generic phrases in your UI like "Track your progress" or "Meet your fitness goals," you might write, "Crush your fitness goals with daily tracking!"

However, make sure that you stay true to your brand voice while you incorporate user language into your writing.