Three years ago I deleted everything that didn’t make the cut. A rejected logo mark, a navigation pattern that tested badly, a color palette the founder hated on sight. All of it went into the trash, because that’s what a tidy file looks like. Then a client at a seed-stage fintech called Harlow asked me a question that changed how I work: “What else did you try?” I had nothing to show her. That gap cost me a week of redoing exploration I’d already done once.
The page nobody asks for
Now every project file has a page called Dead Ends, pinned right after Cover. It holds the concepts that got cut, labeled with the date and the reason they didn’t survive. Not polished. Not annotated to death. Just enough context so a future version of me, or a teammate, understands why door number two got closed.
What goes on the page
I’m selective about what earns a spot. Minor variations on a button radius don’t belong there. What does belong: full directional explorations that took real hours, anything a client explicitly rejected, and any idea I personally loved but couldn’t defend with data or logic.
- A screenshot of the concept at the stage it was abandoned
- One line on why it stopped (tested poorly, client veto, technical constraint, gut call)
- The date, so the timeline of decisions stays honest
- A link to any research or feedback that triggered the cut
Why this matters more than the final file
Clients remember decisions, not files. Six months after launch, someone from the original team will ask why the dashboard doesn’t have a dark mode toggle, or why the onboarding is five steps instead of three. Without a record, you’re reconstructing memory from scratch, usually under pressure, usually in a meeting where you look unprepared.
A real example
For Harlow’s onboarding flow, I built eleven screens around a conversational, chat-style setup before scrapping the whole direction. It felt modern and I was proud of it. But it added forty seconds to average completion time in testing, and the founder’s instinct, that users wanted speed over personality, turned out to be right. That failed direction is still sitting on the Dead Ends page of that file. When a new designer joined the team eight months later and suggested almost the identical idea, I didn’t have to argue from memory. I just sent the page.
That one link ended a debate that could have eaten a sprint.
A decision without a visible alternative looks like an opinion. A decision next to the thing it replaced looks like reasoning.
How it changes client conversations
Showing the discarded work shifts the dynamic in the room. Instead of presenting one polished option and defending it, I’m narrating a process. Clients stop asking “why this?” in a skeptical tone and start asking “why not that?” in a curious one. It’s a small linguistic shift but it changes who’s doing the convincing.
It also slows down bad feedback
When someone asks for a direction I already tried and dropped, I can pull it up instead of re-litigating it live. I did this with a retail client last spring who wanted to revisit a mega-menu navigation we’d killed in week two. Seeing the actual mockup, plus the heatmap data showing people never scrolled past the third category, ended the conversation in four minutes instead of a forty-minute meeting next week.
The discipline of not deleting
The hard part isn’t making the page. It’s resisting the urge to tidy up after yourself, which is a habit most designers build early and never question. Deleting feels like progress. Keeping the mess feels like admitting you wasted time. Neither feeling is accurate. The mess is the evidence that you actually explored the problem instead of jumping at the first workable idea.
I started timestamping my dead ends partly for myself. Looking back at a file from eighteen months ago, I can see how my first instinct on a project has shifted. I used to default to dense, information-heavy layouts. Now my first sketch is almost always sparser than it needs to be, and I add density back in only when testing demands it. That’s a real shift in judgment, and I wouldn’t have noticed it without the paper trail.
What it costs you
This isn’t free. Dead Ends pages take time to maintain and they make files heavier. On a recent project for a wellness app called Linden, the page grew to eighteen entries by the time we shipped. Some clients skim past it entirely. A few have asked me to remove it before a handoff, worried it makes the process look indecisive rather than thorough.
I push back gently when that happens. Indecision is flailing without direction. What’s on that page is the opposite: a record of hypotheses tested and ruled out, which is closer to how good engineering teams already work. Nobody asks a backend developer to hide the architecture they didn’t choose.
Starting your own version
You don’t need a special template. A blank frame and a habit of dragging things there instead of deleting them is enough. The only rule I’d insist on is the one-line reason. Without it, the page is just a graveyard. With it, it’s a map of every decision that led to the thing you actually shipped.
