Skillmark

Overview

Skillmark started as an idea about twelve years ago, when I was an art student surrounded by fantastic artists. I wanted to get to where they were. Learning about the research behind the 10,000-hour rule changed how I saw that goal: skill wasn't magic you were born with. It was time, put in consistently. Ever since, I've tried to track my practice and push myself.

I tried a lot of time-tracking apps along the way, and they kept failing me in the same ways. They were hard to learn and complicated to use. Worse, they left me feeling discouraged. They showed the gap between where I was and where I wanted to be, and never celebrated the work I had already done.

One had a forecast that tracked every day against you. Miss a week and it would tell you, in effect, that you'd die before reaching your goal. A great week afterward barely moved it; every missed day was dead weight it never forgave. I wanted something that did the opposite.

I led the product and its design: strategy, UX, visual identity, content, and monetization. I built it in close collaboration with AI coding tools, directing the technical side as well.

The core principle came straight from those old apps: show people how far they've come before showing how far they have to go. Skillmark turns hours into levels and ranks, from Initiate at 20 hours through Journeyman, Artisan, and Master to Grandmaster at 25,000. Each rank has its own crest, and reaching a new level is a deliberate moment rather than a toast notification.

The early ranks come quickly so a new hobby feels rewarding in its first weeks. The later ones stretch across years, so the ladder still means something to someone who has practiced for a decade.

The first crests were placeholders the AI coding tools drew in code: thin line icons on identical shields, with Master and Grandmaster nearly indistinguishable at icon size. I used quick AI image generation as a direction check, then drew the final set myself as single-color vector artwork, with heavier strokes for small sizes, a distinct silhouette for every rank, and ornament that arrives only as the ranks climb: wings at Artisan, then crowns and rays at Master and Grandmaster.

Set a target and Skillmark turns it into a pace and a projected finish date. This was my direct answer to the forecast that never forgave a missed week. Skillmark weights recent practice most heavily, so a good week moves the projection this week. A slow stretch doesn't become a permanent verdict.

Progress is framed as ahead or behind your own plan, never compared with anyone else. There is no feed, no followers, and no public activity.

Every hobby in Skillmark breaks down into five to eight subskills. I held them to a strict definition: concrete and practicable rather than nebulous, covering the full range of what contributes to the skill, with as little overlap as possible.

That structure does different work for different people. For newer hobbyists, it points attention toward parts of the craft they didn't know mattered, instead of the one part they can see. For experienced ones, the balance view shows where to push to become a more well-rounded practitioner.

The biggest feedback came before launch, during internal and closed testing. Early builds let you set up skills and log time, but they didn't guide you through setup or explain the concepts in play. My mother-in-law loved the idea, but the technical first run was a roadblock to actually using it. Once set up, the app was simple; getting there wasn't.

I rebuilt the first run as a guided setup that personalizes the app and walks you through creating your first skill. It uses progressive disclosure throughout: ask only for what's needed, then show only the result that follows from it.

The hardest part was helping experienced people account for years of practice they never tracked. Two ideas made it manageable:

• Phases. Describe your routine as months or years at a certain number of hours per week. If it changed, add another phase. Skillmark adds it all up.
• Confidence sliders.Rate how capable you feel in each subskill from 1 to 10, and Skillmark distributes your past hours in proportion, on the reasoning that people spend more time on what they're better at. It gives everyone an honest starting point to build from.

The guided setup has been the best-received part of the app with early users.

Early versions organized the home screen around a segmented control: Mastery, Forecasts, and History, each with its own set of cards. The intent was a broad view of each type of information. In practice it became a rug to sweep features under.

People think about one skill at a time, not one data type at a time. Splitting a skill's story across three tabs meant going back to the top to find what else was there, and it hid much of what the app could offer. The redesign brings a skill's progress together, so it's visible without hunting for it.

Tracking time answers “how much?” but not “what should I work on?” My first answer was a learning-resource search that ran a fresh web search every time, targeting subskills, price points, and formats. I loved the idea, but with the time I had to polish it, it stayed unreliable and too often returned poor resources.

I replaced it with a curated catalog: courses, videos, and references for every hobby, ordered from beginner foundations through advanced practice and filtered by subskill. People can browse it all, explore ideas, and email suggestions that get curated into the set. It's less magical, and far more trustworthy.

I didn't want another app that asks for money every month for a small benefit, or another password to juggle and recover. Skillmark doesn't need either. There is no account; data stays on the device, and portable backup files mean nothing moves unless you move it.

A subscription is fair when a product carries real ongoing cost, like cloud storage. A productivity tool that is built once shouldn't be paid for forever. So core tracking is free, not a trial; ads are banners, not pop-ups; and the upgrade is a one-time $5.99 purchase. There will always be a free local version. I can live with a little less revenue to be good to the people using it.

Much of practice happens with the app closed: phone face down on a music stand, or locked in a pocket on a run. Timers keep running and alert correctly when the app is not open, reminders are scheduled natively on each platform, and a stopwatch widget and an iOS Live Activity keep the session visible without unlocking the phone.

I started out unfamiliar with development, so I played the roles I knew: product manager and art director. The AI built what I asked for; I coached it, corrected it, and set the strategy. Once the app worked, I used it every day.

From there I took more direct control: I drew my own assets, refined the visuals in Figma, and used Figma's MCP server to feed those refinements back to the AI. That became the start of a shared design system of components, colors, and tokens.

The recurring challenge was drift: a component dialed in during one round would be ignored by the next AI-built feature. Having the AI document design and system decisions as they were made reduced that, though new colors, progress bars, and charts still need extra attention to stay inside the system.

The hardest new territory was getting into the stores and setting up monetization: business rules, data privacy disclosures, and accurately translating what the ad platforms required into what each app store asks.

Along the way, those lessons became guardrails the AI works within:

• One shared product codebase for iPhone, iPad, and Android, with platform-specific work kept in separate native layers
• Documented design and system decisions for every new feature to follow
• Written architecture rules, including never changing the saved-data format without a tested migration
• Automated tests and checks that run on every change
• A documented release process and checklist for each store

Skillmark is live and early. The first people I shared it with turned out to be like-minded: they wanted to track progress on something that mattered to them, and they came with plenty of ideas. My focus now is growth: the store listing, refining the first run, and finding the communities where people already track their practice.

The biggest thing I take from it: used well, AI lets one designer take a product from idea to both app stores. It covers knowledge gaps with fast research and frees up time for the parts of the work where my judgment matters most.