<- Back to all posts

The most important part of your design is actually the text

Jessica Ouyang
July 9, 2020

Design is about constantly closing the gap between what users expect from your product and what it actually does. Product teams try to close this chasm every day, whether by user testing, tracking metrics and behavior, or seeking feedback. In attempts to bridge this user experience communication gap, teams often forget the most important thing — the text.

Product copy, or the text found on user interfaces, is the most under-leveraged part of product development today. Here are a few reasons why.

Editing copy is your highest ROI lever

Unlike almost anything else in product development, editing copy doesn’t require any engineering or design lift. Working on copy is about maximizing the utility of your existing feature set.

In many cases, the cheapest and most practical solution to a problem is well-written copy. Rather than committing a team’s time to building an entirely new feature, editing the copy goes a long way in user retention and engagement.

Whether it’s a clearer error state that helps users self-correct or an empty state that helps users get started, editing text can directly affect metrics. [1] It’s easy for the conversation around copy to revolve around marketing copy (thinking about text on a landing page or a marketing email and whether or not it converts), but this conversation around the importance of text needs to shift to product as well.

Of course, there’s no substitute for poor design, but bad copy can make even the best designs ineffective.

Product copy is your most direct line of communication to users

As John Maeda expressed with his 2017 Design in Tech report, “Words are really important because the graphics don’t make sense sometimes.”

Product copy makes expected user behavior explicit. It’s your chance to communicate intent. In being what your users actually read, the text anchors the user experience and bridges gaps in expectation. As UX writer Ryan Cordell points out:

“Microcopy is how your product communicates. It provides answers, feedback, comfort, guidance, encouragement and more. If we fail to design the words we use, this still communicates something to the user. It says we can’t or don’t want to help. It says we don’t care as much as we should. And that’s catastrophic for the customer experience.” [2]

Ultimately in building a product, you’re optimizing for user understanding — to surface the value of the features that your team spent so long to build.

Text is what guides users when they venture off the happy path

It’s often easy for designers and PM’s to optimize for the happy path: the user experience that most closely aligns with their own expectations of how the product should be used. However, we all know that expectations often exist in isolation from how people actually use your product — and much of the burden of steering users back on track relies on the text.

Copy serves as your product’s guardrails. Whether it’s an error message, a new user experience, or an empty state, it’s often text that guides users when they don’t know what to expect. In a study conducted by Microsoft Azure’s design team, user test errors decreased by 44% after copy changes alone. [3]

It’s in these moments of mismatched expectations that clarity is most critical. This is especially true as the role of software expands. From health-tech to purchasing houses and cars, the sole responsibility of guiding people through complex and important decisions is now carried by the digital experience.

Often times, however, because these non-ideal users journeys don’t get designed, users are greeted with generic text written ad-hoc by the last person that touches it.

Copy exists from end to end

Almost every single step of the product development process works in verticals — designers work in design tools and then pass designs off to developers that work in devtools. The same verticalization of tooling and work applies to almost every other step of building a software product with a team.

The exception to this verticalization of product development is the copy, which exists from the very beginning (specs, drafts) all the way to the very end (in the final product) and is touched by everyone from design to engineering to legal to marketing.

Rather than existing as one-offs, the product text works in a system to build trust for users. Like the narrator in a story, copy sets the context and defines the goals. [4] Thinking about copy early means being intentional about the information you convey to users — and drives design decisions like progressive disclosure and information architecture.


Let’s be clear: writing good copy is hard. It might not be glanceably flashy the same way a Dribbble post might be. Text often “looks” the same in a design at first- and second-pass, even after edits. Reworking copy can be painstaking — and requires an in-depth understanding of both the product and the user.

But as we want to build and design products that shape how people interact with the world, we shouldn’t overlook the obvious when communicating: the words.



[0] In full disclosure, our belief that text should be first-class in building product has led Jolena and I to building Ditto. 🤗

[1] Even really simple copy changes change produce enormous ROI. One example of many is the 17% increase in engagement when Google Travel replaced “Book a Room” with “Check Availability”.

[2] You can find more great excerpts from Ryan’s post on communication theory and microcopy here.

[3] In addition to decreased user test errors, the metrics-driven insights found by Kylie Hansen‘s team impacted NPS, customer retention, active users, and more. She goes over some of it here.

[4a] A scenario that often requires high trust is a monetary transaction. Increasing transparency by clearly surfacing what to expect is something that both Airbnb and Amazon do well.

[4b] Over coffee, John Saito once mentioned to us that the product of Dropbox itself had more words than a Harry Potter novel!

Success! 🥳 Look forward to Ditto updates in your inbox.
Oh no — something went wrong while submitting the form. Please try again!