Would you write an email without knowing what it is for or who the recipient is? Well, if you’re not testing your content this is exactly what you’re doing, along with deleting the answer before reading it. If you want to start testing content or get better at it, this article is for you.
Who is responsible for content testing?
Before talking about content testing, it is crucial to understand that it needs to be part of a whole and that whole is our content strategy. I like to think that developing a content strategy is telling a story. And every good story needs:
- A theme and point of view which can be understood as the brand guidelines and the marketing goals;
- A setting, which is the time, location, and atmosphere that we may read as the market research;
- Characters and their motivations, as our audience personas and their empathy maps;
- A plot that defines the sequence of events and its logic, thus our roadmap and framework;
- A resolution of events, hence our measurement plan to get insights on how to evolve our strategy.
If you want your content to provide value to your audience and achieve business results, then you are responsible for making content testing one of the main characters of your story.
What is content testing?
Content testing is a discipline that brings together research methods to assess if the written content is aligned with the predefined objectives and ultimately adds value to its target audience. Text is the most direct way to translate the brand’s voice into a digital product, it’s how the interface talks with users and how it becomes human. Testing content is a way of prototyping that conversation.
Why should we test content?
A prototype is a tool that helps us reduce risk and serves as the basis for iterating and evolving. So, prototyping and testing content reduces uncertainty and avoids putting effort into content that doesn’t work.
Besides measuring whether the content is findable, readable, and understandable – which are basic usability matters – our goals may also be to educate, inform, entertain, inspire, convince… and these can also be tested. We can measure satisfaction (attraction/value), effectiveness (tasks completed), efficiency (in terms of relevance, findability, and accessibility), reading comprehension, voice and tone alignment… and the list could go on with any other goal we wish to pursue through content.
When to test?
If we google “content testing” we’ll discover that most of the entries are about how to validate content once it gets published. But now that we know that the digital content creation process finishes when users interact with our text, wouldn’t it be better to involve users from the very beginning? What about co-creating with them? Could we then iterate and evolve our prototypes based on these co-creations? Yes, yes, and yes.
Testing from the early stages of a project’s development goes hand in hand with a content-first approach. Going content-first means prototyping with beta or draft texts, betting on their continuous improvement, and allowing to spot errors and optimizations earlier, with less expensive corrections. In this way, design, content, and functionality are developed simultaneously, placing the user’s experience at the center (the antithesis of this approach is what is known as “lorem ipsum”, which is designing and prototyping with fictional texts).
So, when to test? Short answer: before, during, and after the content release.
How to test content? (and where we can do it)
- Define the test’s objective.
It must be aligned with the content goals, which in turn respond to the business’s main purpose. The objective can be quantitative or qualitative, but it should always be measurable.
- Plan: sample + participants + method.
Once we know what we want to test, we need to ask ourselves the following questions that will help us define how to carry out the testing: which parts of the content we’ll test, with whom, and using which method.
Probably, we won’t be able to test all the content, so we need to choose the sample that will most likely give us the information we need. Also, content is not good or bad, but good or bad for someone. That’s why selecting the right participants is so important: they must be real or potential users aligned with our content personas. Finally, time and budget also play hard in this instance.
- Run the test and document its results.
- Analyze and learn: interpret the results looking for patterns and insights to define action points.
Content testing is a relatively new discipline so it encourages us to be creative when deciding how to test. As long as we test with rigor, everything is allowed! Besides surveys, interviews, focus groups, workshops, readability and usability tests, some of the methods that we can use are:
- Tell a friend: We ask participants to “tell a friend” what a section of the product is about. Analyzing these stories will provide us with data on intelligibility and wording.
- Cloze test: We erase each n word of a text (usually the 5/6th.), leaving a blank space for participants to fill in with the term they think would match. Analyzing hits/misses/recurrence will help us measure intelligibility and adjust copies.
- A/B/n test: We present two or more versions of a piece of content to collect quantitative data of which of them generates better results.
- Highlighting test: We pick one of our texts and ask participants to highlight what is most valuable to them (according to the content objective) with one color, the least valuable with another color, and what they found indifferent with a third one.
- Card sorting: This exercise invites participants to categorize content classifying elements. We then search for patterns in the way participants group and associate cards.
- Tree-testing: It evaluates the structure and hierarchy of content elements, if the wording or tagging is effective, and, therefore if the findability ratio should be improved.
- Wireframe session: It involves tasks that need interacting with the text, allowing to test the content experience as a whole: for example, browsing a web or going through specific flows (find a content, understand it, evaluate it).
- Happy or not: Test in context to get feedback from users at the time of content consumption.
- Eye-tracking/heatmaps: Eye-tracking studies track where the user is looking when interacting with our content. We can carry out qualitative (gaze plots and gaze replays) and quantitative (heatmaps) studies.
On where to test, there are lots of digital tools to do so. Some of them combine several types of testing while others specialize in one type. You should find the one that best suits your needs and budget.
What’s next for content testing?
The latest challenges in content testing aim to use artificial intelligence to develop better ways to test engagement. If we want to evaluate the emotional commitment to our content, we need to be able to measure not only reactions but also feelings.
In the future, content testing will involve a combination of technologies to collect physiological and biometric data (such as eye-tracking, facial expression coding, electrodermal activity -EDA-, galvanic skin response -GSR-, or electroencephalography). It will also include AI software to help interpret that data with learnings from neurology, biology, psychology, and sociology.
