AutoDesk · UX/UI Lead Designer
Pencil experience
Cross-platform, mobile-first solutions for architects and engineers.
No launch analytics; instrumentation was a known gap. See Impact.
Goal
Enable users to leverage Apple Pencil and stylus in AutoCAD mobile: natural sketching, precise editing and seamless workflows, while keeping cross-platform consistency.
The challenge
Users expected Apple Pencil to behave like a real pencil: instant drawing, intuitive gestures, accuracy. Many struggled switching between finger and pencil, lacked shortcut support, and wanted better control of tools.
Discovery
5 inputs01. Target users · persona
Architects and engineers on the go, sketching between site and studio.
On the go, they want to sketch, edit and measure quickly with tools that are easy to access and feel natural. The ideal setup supports freehand drawing, smart navigation, and shortcuts that match how they already think about using a pencil.
User interviews · UserTesting.com
02. Customer support & reviews
The App Store was already telling us what was broken.
Frequent crashes, an inability to use the Pencil for drawing, and mismatched cursor-versus-tip accuracy. These issues made the app unreliable for daily work.
App Store reviews & direct feedback
03. Benchmarking
Strong sketch tools everywhere, weak hybrid finger–pencil workflows.
We explored how stylus support was handled across leading apps, noting strong sketch tools but weak support for hybrid finger–pencil workflows.
Competitive benchmarking of CAD & drawing apps
04. Research & observation
Pencil means create; finger means navigate. Users had already decided that.
- Users expected pencil = creation, finger = navigation.
- Users requested customizable double-tap shortcuts.
- Freehand sketching was the most expected feature.
Observation sessions with early users
05. Best practices · HCI literature
A pencil must mark instantly. Latency reads as a broken tool, not a slow one.
Designing for stylus interaction means matching users' mental model of physical tools: a pencil must mark instantly, with no delays or mode switches. Effective Pencil-first design, per HCI research and Apple's HIG, emphasizes zero-latency feedback, modeless interaction, smooth finger–Pencil transitions, and quick recovery through undo and correction gestures.
iOS guidelines, UX laws & HCI literature
What discovery told us
- Users expect pencil = draw, finger = navigate.
- Freehand sketching is the most requested feature.
- Reviews highlight missing smart-pen behaviour and lag.
- Benchmarks show strong sketch tools but weak hybrid workflows.
- Users want customizable double-tap shortcuts.
- Pencil latency is perceived as a quality failure, not a bug.
What made it hard
- Inconsistent input behaviour breaks users' mental model.
- Too many manual mode switches slow expert workflows.
- Even minor Pencil latency destroys trust.
- Pressure actions lack clear visual feedback.
- Advanced gestures are powerful but hard to discover.
- Cross-platform differences force users to relearn tools.
Design
Flow and 5 featuresUser flow
Pencil detected, tool chosen, stroke committed, with an exit at every step.
Solution & features
Feature 01. Gesture support
Decision: map each gesture to the tool it implies
Tap, drag, press and hover were mapped to the right tools, so the input itself selects the intent rather than forcing a mode change first.
Feature 02. Freehand tool
Decision: ship the thing everyone asked for first
A cross-platform drawing tool: the most expected and most natural behaviour, and the feature that made the Pencil worth holding.
Feature 03. Pencil mode
Decision: block the palm, keep the pencil Experiment
Unwanted finger touches were blocked while the Pencil was active, improving accuracy. Shipped as an experiment rather than a default, since suppressing touch input is a decision users need to be able to reverse.
Feature 04. Double-tap shortcuts
Decision: let the user choose what the tap does
A customizable double-tap for quick actions: undo, eraser, or a tool switch. Users had asked for this specifically, and a fixed mapping would have suited only some of them.
Feature 05. Pressure sensitivity
Decision: pressure drives precision, not just line weight
Pressure input was supported for accurate pins, points and line weight, so the same gesture could place something precisely or draw something expressively.
Impact & next
Verified before releaseCustomer review and observation
Verified by hand, because the analytics weren't there to verify it for us.
We verified pressure sensitivity, shape-recognition accuracy, and seamless switching between finger and Pencil before release.
QA sessions, internal review meetings
Success metrics defined
- Reduction in manual mode switching per session.
- Increase in Pencil-initiated drawing actions.
- Time to first successful freehand sketch.
- Decrease in undo-after-commit events.
- Increase in dimension and annotation usage.
- App Store review sentiment related to Apple Pencil.
Trade-offs accepted
- Faster drawing vs. accidental commits.
- Modeless flow vs. loss of system visibility.
- Clean UI vs. hidden gestures.
- Desktop parity vs. native Pencil behaviour.
- Custom shortcuts vs. onboarding complexity.
Next
Refine pressure thresholds through usability testing to reduce accidental commits, improve discoverability with subtle in-context hints, and keep validating stability through crash monitoring. Since analytics instrumentation was difficult here, introduce lightweight events gradually: Pencil detected, first freehand stroke, pressure commit used, undo after commit, and the finger-versus-Pencil interaction ratio, to compare behaviours and measure long-term adoption.