.png)
You're racing toward launch. The designs look perfect. Then someone asks, "What about the content?"—a question that far too often gets asked too late in the process. Suddenly, the sky seems darker than before, and that perfect timeline starts to crack. Whether you're a designer working without dedicated writing support or a content designer pulled in at the eleventh hour, here's how to keep things afloat amidst the storm.
I wouldn't call myself a skeptic, but it's not by chance that I became a content designer. We have a certain proclivity to question. A way of naturally breaking things down to their atomic parts and asking how and why they got there in the first place. When I learn about a new process or tool, my mind immediately looks for the gaps. The missing pieces. There's a healthy sense of mistrust in the ways I approach work that, for the most part, has served me well in this field.
It's also why, when I read idealized takes on ways of working within content design, I swim through waves of doubt and envy–each wave crashing onto me as I dive further and deeper into the content-related think piece–looking for the "aha! I see" moment where their optimism seems justified (and grounded in reality).
They're talking about how things should be, not how they actually *are*
Their practice has been around for decades. They've had time to figure things out.
Oh, things aren’t broken yet because they’re the first content design hire.
And so this is for the designers, product and content, who read or listen with their mouths agape, in disbelief, to stories of near-perfect process across design teams and smooth delivery with developers.
The prospect of writing product content is more daunting than ever before. That's because there has been, of late, an inrush of generative AI tools that promise clear, concise, and conversational outputs with just a few easy prompts.
Prompt: Write 20 iterations of clear, concise copy for a button that takes members to a credit card marketplace.
LLM: Here are 20 iterations for a credit card marketplace button, arranged from most direct to more benefit-focused.
As a result, this has led to a degradation of process and, in turn, quality writing outcomes. And so the collective stance on what “good” content is has been replaced by “good enough”. But.. of course it did. This is the logical conclusion of prioritizing velocity over quality and instant gratification over meaningful outcomes.
Now, given its uncanny and lightning fast success, generative AI may very well close (what is now) the quality gap between writers and LLMs to a very small margin in a few years. But no matter when that day comes, it's worth grasping the science and art of communication in ways technology will never be able to replicate. That said, no amount of resisting AI will do you much good at this point, so it's best to accept the state of things and work to maintain control in the ways we can.
So, product designers, let's take it from the top. You've been asked to write the content in your designs and the blank frames are waiting. Here's your gameplan.
You don't need to be an experienced writer to produce something usable. What you need is a solid foundation of research.
Content methodologies like tree testing and card sorting are miracle workers because they'll give you the scoop on your audience's mental model(s), how users perceive terminology, and, ultimately, where they expect information to be (i.e., information architecture). The tests themselves don't require too much prep work, and they'll do wonders for findability.
When you get to the actual writing part, give yourself permission to be bad. To write a first draft that should not see the light of day. But stay the course–keep rewriting, re-editing, and redoing what you've done so far–because there aren't any shortcuts, and it'll take time, but you'll have something prime for testing.
Storytelling is a skill every designer should hone since the overwhelming medium of choice these days is Figma. So, what makes a good story? Well, for starters, a consistent narrative arc. Building text components to ensure the story stays consistent is a stellar way to do this. Buttons, information disclosures, headers–anything can be componentized to save yourself the task of updating every instance after a change or slight rewording. Lucky for you, Ditto allows for component creation and shows the version history, which captures how the content changes over time.
Speaking from personal experience, I used Ditto in this very capacity to archive old methods, like copy decks, when we launched our student loan experience during the pandemic. As a sole designer among a sea of cross-functional partners, there were 1000+ comments from our legal counsel in one Figma file–and if there's one thing other teams care about, it's context. Using Ditto, I was able to see, show, and track all of the edits by all of the contributors. The ability to have conversations in Figma without oscillating between other tools and screens was also a nice touch.
It's hard to be a content designer these days. It just is. We receive feedback all the time about how we can't take direction or take exclusion too personally and that we shouldn't be too attached to our work when it's ignored or changed. And as much as that well-intentioned advice almost always rubs me the wrong way, I use it to make me better. I hope the same goes for you after reading this.
I wouldn't be surprised if the phrase most used in my first few years at Credit Karma was "short beats good" because it's one of my favorite principles in our content design system. It's not novel or pithy but a helpful reminder that people don't read word-for-word on the web or in mobile apps. Instead, they prefer short, concise text and don't have time for much else.
Thanks to APIs, this also benefits product designers and engineers.
We saved our developers time and headaches by using dev mode to assign unique text IDs to our strings. We didn't have a centralized way of doing this in the past, so typos, broken links, and old content would go unfixed because of the required level of effort.
Ditto didn't just help us manage our content better going forward; it fixed our past grievances by also assigning the unique IDs to legacy strings as well. It magicked our strings findable, saved dozens of hours in QA each quarter, and deepened the sense of ownership in our design process.
Ultimately, designers want to know whether their designs pass the sniff test.
If they've done the work to take an earnest pass at documenting context and giving you enough to go off of, or there's nothing destructive in the frames, grit your teeth and give the go-ahead.
But in the same breath, ask.
Ask for the opportunity to come in earlier and fix that convoluted login and account recovery flow (you know from your research that it's responsible for a lot of bounces). Give them your elevator pitch, and know that reciprocity is more likely on the heels of giving someone else what they want.
The world is heavy today, and there’s plenty of learning for many days to come. Show yourself the grace to take small steps, low-risk chances, and accept that even in a dysfunctional situation, you can still come out on top and feel good about it.
