The Log

Decisions That Mattered

Design decisions look obvious after the fact. They never are in the moment. This is where I document the ones that actually moved something. Not the pixel choices. The product choices. Why we killed a feature users said they wanted. Why we moved a CTA that was performing "fine" and watched signups climb. Why a client's rebrand failed and what the real problem turned out to be.

Each entry is short. One decision. The thinking behind it. What happened next. I keep this because I've learned more from documenting my own decisions than from any design book I've ever read. Real decisions from real engagements. Names and companies are kept anonymous to protect client trust. The thinking is what matters, not the logo.

1. We Removed the Feature Tour

The Situation A product was losing over a third of new users during onboarding. They had a five-step feature tour that explained everything the product could do. Users had to click through five screens before they could touch anything.

The Decision Kill the tour entirely. Drop users directly into the product with one single action prompted on screen.

The Thinking Feature tours optimize for education. Onboarding should optimize for momentum. The goal isn't making users understand your product. It's giving them one small win so fast they want to keep going. Understanding comes after commitment, not before. Nobody reads the manual before driving the car.

What Happened Activation rate increased meaningfully within two weeks. Support tickets didn't go up. Users figured things out because the first action was designed to be obvious. The tour was solving a problem the product should have solved through design.

The Takeaway If your onboarding needs a tour to explain your product, the product has a clarity problem. Fix the product, not the tour.

2. The Homepage That Said Too Much

The Situation A client's homepage had seven sections, three CTAs, a product video, testimonials, a feature grid, and a pricing table. Everything a marketing team could ask for. Bounce rate was over 65%.

The Decision Strip it to one headline, one subline, one CTA, and one visual above the fold. Everything else moved to secondary pages or got cut entirely.

The Thinking A homepage has one job: make the visitor care enough to take one action. Every additional element competes for that attention. Seven sections means seven chances to lose someone. Comprehensive isn't clear. Clear is clear.

What Happened Bounce rate dropped significantly. CTA clicks nearly tripled. Time on page actually went down, but conversion went up. Users didn't need to read more. They needed less noise to decide.

The Takeaway Clarity converts. Comprehensiveness confuses. If your bounce rate is high, the problem usually isn't that you're saying too little. It's that you're saying too much.

3. Why We Didn't Redesign

The Situation A founder came in wanting a full product redesign. The product "looked dated." Users were churning. The assumption was that a fresh visual layer would re-engage them.

The Decision Didn't redesign. Instead, we rebuilt the first five minutes of the user experience and rewrote every piece of copy in the core flow.

The Thinking The product wasn't ugly. It was confusing. Users weren't leaving because of how things looked. They were leaving because they couldn't figure out why the product mattered to them. A visual redesign would have been expensive camouflage over a messaging problem. You don't repaint a house when the floor plan doesn't work.

What Happened Churn dropped noticeably over the following quarter. The product still looked the same visually. Nobody complained about the aesthetics once the experience made sense.

The Takeaway Most design problems aren't visual problems. They're clarity problems wearing a visual costume. Before you redesign anything, ask: do users understand what this does and why they should care? If not, the redesign won't fix what's broken.

4. Delaying the Commitment Step

   
   

The Situation A DeFi product required wallet connection as the first step. A huge percentage of new visitors were bouncing at that screen before ever seeing the product.

The Decision Let users explore the interface first. See potential returns. Interact with sample data. Understand the value. Then prompt the wallet connect.

The Thinking Wallet connect is a trust moment. You're asking someone to give access to their financial identity before they understand what you do. That's asking for commitment before demonstrating value. No different from a SaaS product demanding a credit card on the first screen. Sequence matters. Trust is earned in steps, not demanded in one.

What Happened Connection rates roughly doubled. Users who connected after exploring showed significantly better retention than those who connected immediately under the old flow. Giving people context before commitment didn't reduce urgency. It increased confidence.

The Takeaway Earn trust before asking for commitment. The sequence of your experience matters as much as the design of any individual screen. This applies to wallet connect, signups, payment flows, and anything where you're asking the user to give you something.

5. The Button Nobody Clicked

The Situation An AI product had a "Try it" button on their landing page. Traffic was healthy but conversion was flat. People were reading but not clicking.

The Decision Changed the CTA copy from "Try it" to "See what it finds for you." Added a single input field directly on the homepage so visitors could paste a URL and see results without creating an account.

The Thinking "Try it" is a request. It asks the visitor for a favor. "See what it finds for you" is a promise. It offers a result. And making the product work on the homepage before signup removes the biggest unstated objection: "Is this worth my time?" Let them answer that with experience, not faith.

What Happened Signup conversion increased substantially. Most users who saw results on the homepage created accounts in the same session. The homepage went from a marketing page to a product experience.

The Takeaway Don't ask people to try your product. Show them what it does to their problem, right now, on the page they're already on. Reduce the gap between curiosity and proof to zero.

6. Killing the Feature the Team Loved

   
   

The Situation A product team had spent two months building an advanced analytics dashboard. It was well-designed. The engineering was solid. The team was proud of it. Usage data after launch showed almost nobody was using it.

The Decision Kill it. Remove it from the navigation entirely. Redirect the team's energy toward improving the three features that 80% of users actually relied on daily.

The Thinking Sunk cost is the enemy of good product decisions. The dashboard was built on an assumption about what users wanted, and the assumption was wrong. Keeping it alive would mean maintaining code nobody used, adding cognitive load to the navigation, and sending the message to the team that shipping matters more than impact. The hardest part wasn't the decision. It was the conversation.

What Happened Nobody noticed it was gone. Literally zero support tickets. The three features that got the redirected attention all improved measurably over the next quarter. The team initially resisted, then later cited it as one of the best product decisions of the year.

The Takeaway What you remove from a product matters as much as what you add. If a feature isn't being used, it isn't neutral. It's actively making the product harder to navigate, maintain, and understand. Kill it. Put the energy where it compounds.

7.The Brand That Tried to Be Everything

The Situation A startup had repositioned three times in eighteen months. Each time, they updated the visual identity to match the new direction. By the time they reached out, they had a Frankenstein brand: colors from one era, typography from another, illustrations that matched nothing, and a voice that shifted on every page.

The Decision Stop. Don't redesign again. Instead, write down who you actually are in one paragraph. Then strip everything back and rebuild from that single truth.

The Thinking Brand inconsistency isn't a design problem. It's an identity problem. If you don't know who you are, your visual language will always feel scattered because it's trying to represent something that hasn't been decided yet. The brand work had to start with words, not visuals. Once the identity was clear, the design decisions became obvious.

What Happened The rebrand took less time than any of the previous three attempts. Because the identity was defined first, every design decision had a filter to pass through. The result felt cohesive in a way their previous attempts never did. More importantly, the team could make brand decisions independently because they understood the principle, not just the style guide.

The Takeaway If your brand feels inconsistent, the problem isn't your designer. The problem is your team hasn't agreed on who you are. Solve the identity first. The design will follow.

8. Speed as a Design Decision

The Situation A product was experiencing declining engagement. Features were shipping. Nothing was technically broken. The team was focused on adding capabilities. Meanwhile, the app had gotten noticeably slower over six months of feature additions.

The Decision Freeze feature development for a full sprint. Dedicate the entire team to performance. Cut page load times, optimize animations, reduce unnecessary API calls.

The Thinking Speed is a design decision. Every 100 milliseconds of delay creates friction. Over thousands of interactions, slow products train users to hesitate. Fast products train users to trust. The product hadn't gotten worse in terms of features. It had gotten worse in terms of feel. And feel is what keeps people coming back.

What Happened Engagement metrics recovered within weeks of the performance sprint. No new features were added. The product just felt better. Users described it as "snappier" and "more responsive" in feedback, even though the interface looked identical.

The Takeaway Before you build new features to solve an engagement problem, check your speed. If the product feels slow, nothing you add on top will fix the underlying friction. Performance is invisible design. And invisible design is often the most important kind.

9. The Pricing Page Nobody Could Parse

The Situation A SaaS product had three pricing tiers. Feature comparison table with forty rows. Check marks and X marks everywhere. Users were spending time on the pricing page but not converting. The team assumed they needed a fourth tier or a free trial.

The Decision Redesign the pricing page. Not the pricing. The page. Cut the comparison table from forty rows to eight. Highlight the one feature that differentiated each tier. Add a clear recommendation for each user type.

The Thinking Choice overload kills conversion. Forty rows of feature comparisons don't help users decide. They help users procrastinate. The user doesn't want to compare features. They want to know: which plan is right for me? The pricing page should answer that question in ten seconds, not ten minutes.

What Happened Conversion on the pricing page increased significantly. Interestingly, more users chose the mid-tier plan, which was the highest-margin option. When you reduce choice complexity, people default to the recommended path. That's not manipulation. That's good design.

The Takeaway If your pricing page isn't converting, the problem usually isn't your pricing. It's the way you're presenting the decision. Make it easy to choose. Highlight the differences that matter. Recommend a path. People don't want more options. They want more clarity.

10. . When the Design System Became Shelfware

The Situation A team had spent four months building a design system. Beautiful documentation. Comprehensive component library. Detailed usage guidelines. Six months after launch, adoption was below 30%. Most of the team was still building components from scratch.

The Decision Stop adding to the system. Instead, embed the system into the team's existing workflow. Set up pairing sessions between system maintainers and product designers. Make adoption the metric, not coverage.

The Thinking A design system that nobody uses is just a Figma file. The team had optimized for completeness when they should have optimized for adoption. The system was built in isolation, then handed to the team and expected to stick. That never works. Adoption is a people problem, not a documentation problem.

What Happened Over the following quarter, adoption climbed to over 70%. The key shift wasn't the system itself. It was the pairing sessions. Designers who sat with system maintainers for thirty minutes understood the system better than any documentation could explain. The investment shifted from building more components to making existing components easier to use.

The Takeaway Design systems fail at adoption, not at creation. If your system isn't being used, the answer isn't more documentation. It's more proximity between the system team and the product teams. Build less. Embed more.
Making decentralized apps intuitive