Write better microcopy with conversation mining [Case study #1]
Let's go word-mining to develop a voice that speaks the user's words.
Hello 👋🏽
Welcome to A UX Writer's Journey.
This is the first post in the case study/scenario analysis series. Posts in this series will use real or hypothetical examples to explore methods, tools, and processes and their application to UX content. I hope you enjoy this part of the journey with me.
PS: I'll use service as an umbrella term for products and services in this post. That way, I don't have to repeat both in the same sentence.
With that in mind, let’s step on the gas, shall we?
Today, I’ll explore conversation mining, widely used in fields from artificial neural language processing to customer service. In this hypothetical case study, however, I’m applying it to UX content (learn more from this article by the UX Writing Hub).
Conversation mining as a UX writing tool is used to gather insight from user-generated content (UGC) to write microcopies that resonate with a target audience.
These UGCs are found in online forums, comment sections, product reviews, and other social spaces where people freely talk about their experience with a service. Users can express positive, negative, neutral, optimistic, excited, or sad feelings.
Gleaning from these conversations helps UX writers develop a brand voice more attuned to their users. And if it resonates, the experience feels familiar because it uses words and phrases the audience already knows. That way, users feel at home while completing tasks easily.
The Case
Readr: beyond reading
Why Readr?
Readr is a book summary app that helps users take strategic action after reading a book. Readr provides non-fiction self-improvement, self-help, and productivity books. Each book has a tailored gamified action plan.
What’s the problem?
Many dream, but few dare. Readr found that the top question on people’s minds is how to translate goals into reality. The following are the highlights from their research:
The problem is not the willingness to take action, but the information overload from the internet and a lack of trusted guidance
Competing apps and platforms are pricey
A lack of step-by-step action plan based on each book
Who are Readr’s target audience
Broad
Anyone who loves reading books, articles, blog posts, online news, and other long-form content online.
Specific
Anyone can read for different reasons, but Readr targets those who read self-improvement, self-help, and productivity books and want to take action on what they’ve read.
Who are Readr up against?
Readr is up against competing platforms like Headway, Deepstach, and Blinkist.
Mining the words
Let’s get our hands dirty. The first thing I did was to get a top-level understanding of Readr’s goals and the user’s goals. Readr’s goal is to get people to take action after reading a book on their platform. To establish their audience’s intent, I used the following keywords on Answer The Public:
Dream to action
Take action after reading
Take action after reading a book
I got exciting insights. Using the first keyword, I focused on what people asked using the ‘how’ question.
Using the second keyword, I focused on all the questions.
The third keyword is even more specific. Besides books, people also read articles, news reports, blog posts, and more. But the focus here is on books.
The ‘how’ questions were punctuated with words like actualize
, achieve
, and take action
. Having established this fact, I used a more specific keyword, ‘take action after reading,’ with revealing results. People want to know how to take action and what to do after reading. Let’s take this further.
The third keyword isn’t dramatically different from the second, but one thing stood out: how people feel after reading a book. Feelings are subjective, but they reveal whether the reader is confident they have gained something. And if they have, it is just natural to want to put what they’ve learned to work.
Digging deeper into social spaces
First port of call: self-improvement book reviews on Amazon. Using the keyword ‘self-improvement books on amazon’ on Bing, I clicked on the first search engine results page (SERP) link. I randomly picked some books and looked through the reviews. I also did the same on Goodreads and found it helpful for this project.
Next, I perused Reddit. The social platform is a fertile ground for conversation mining. Redditors ask questions, express their feelings, and give their honest take on services they have used or plan to use. I explored book clubs and competitor pages on Facebook and Twitter without much help. I searched broadly for book reading clubs but didn’t get much use.
Last, I explored reviews on competitors via the App Store and Play Store. I got insights into what customers were experiencing using those platforms and the words they used to express their feelings.
You can find the data on Airtable here.
Synthesis
The main goal of conversation mining is to understand how users talk about their experience with a service. Sometimes, you get more than that, actual phrases and words that can be used in the UX copies. Other times, you don’t.
Mining words is more than writing; it is a way to get a 360-degree view of users’ fears, frustrations, goals, and wins. So, I didn’t drive too hard because I didn’t get catchy phrases and power words.
There are places I didn’t explore (1) because I had limited knowledge, (2) due to time constraints, and (3) possibly the business goal was not clearly defined.
Below is what best describes what the target audience wants:
Grab actionable insight from a book in a short time
Use the information in a personal way
Use the app wherever and whenever
Access to goal tracker irrespective of subscription plan
Access to quality books in audio (*no-brainer)
Armed with this, let’s craft sample microcopies and web copies for the app onboarding and hero section.
A good way to test out these options is A/B testing, but which of these options would you go for and why? You might even have a different take on this approach. Let me know in the comments section.
There are other considerations to note, such as brand voice and tone, user state of emotion, and context. But conversation mining provides a solid ground to build on.
Other approaches to understanding user goals and lingos come in handy when there are constraints like time and budget. Torrey Podmajersky explains how to use role-play in her book Strategic Writing for UX (pages 40-41). The Microcopy Canvas by Jane Ruffino also helps understand user goals and find the best way to write for emotional state and context.
That’s it for today. Thank you for tagging along on my journey. Next time, I’ll share traits that have been helpful on my journey. Till then, I’ll journey on in cruise mode.😎