Nemetschek Creator — Onboarding

Onboarding that guides first-time users through a new 3D design tool without forcing a fixed path.

Helping first-time users learn a new 3D design tool through suggested tasks, layered help and animation.

Nemetschek Creator is a new browser-based tool for early-stage building design. I owned the design of its onboarding. Instead of a forced product tour, I designed an optional task list, help at several levels, and 12 Lottie animations that teach the core loop: draw a site, draw a building, run an analysis. Creator launched in early access in October 2026.

Year

2026

2026

Role

Sole Product Designer · Onboarding

Sole Product Designer · Onboarding

Work type

Web App

Web App

Project scope

Onboarding UX, User Flows, Interaction Design, Motion Design (Lottie), Content Structure, Developer Handoff

Collaboration

Product Owner, 3–4 developers, QA, Graphisoft learning materials team

Context and my role

Creator helps architects, and people with less 3D experience such as real estate professionals, quickly test large building concepts on a real site. Users draw sites and buildings, check area metrics and compliance targets, and run sun, wind and carbon analyses.

Before the product got the green light, I was one of three designers who shaped its early concept. One idea set us apart from tools like Giraffe and Spacio: Creator should feel game-like. That meant friendly, visual, and easy to start, with lots of illustrations and animations.

Once Creator went into production, I owned the onboarding design end to end. I worked with a Product Owner, 3–4 developers and QA, and with the team that creates learning materials across Graphisoft products.

The problem

New users open a 3D tool with a lot going on: drawing tools, metrics panels and several types of analysis. The onboarding had to serve two very different groups in the same flow:

  • Architects who are comfortable in 3D and don't want to be led by the hand.

  • Users with little 3D experience who need to know where to start.

It also had to match the product's playful tone. A rigid tutorial would have worked against everything we wanted Creator to feel like.

Constraints

  • The product has its own order. A building needs a site, and an analysis needs a building. The onboarding could not hide that, but it shouldn't add rules of its own.

  • One learning approach across products. Content had to fit Graphisoft's wider learning materials, not just this one tool. I also checked that tooltips and panels use the same name for every tool.

  • Easy to build and maintain. Animations had to work for developers without constant designer involvement.

Process and key decisions

1. Guide, don't gate

Most competitors use step-by-step tours that users must finish. My first idea was similar: lock later steps and groups until the earlier ones were done. I dropped it, because it went against what we set out to build. Creator should feel open and playful, not like a course you have to pass.

I also designed a classic product tour that walked new users through each toolbar, step by step, before setting it aside for the task list. The principle we kept coming back to: a guide, not a tutorial.

So every step is available from the start. Users can open any step, read the instructions or watch the video before touching the tool. The only order is the product's own: you still need a site before a building.

The first task, "Create your first project", is already ticked off when the list appears. Users start with a small win instead of an empty checklist.

2. Deciding what to teach first

I worked with Graphisoft's learning materials team to decide which steps belong in the list and how to group them. We didn't want one path built around a single use case. Instead, we grouped tasks around tools that build on each other without locking users in.

The first group covers the core loop: draw a site, draw a building, run an analysis. It's the fastest way for a new user to see what Creator is for. Later groups go deeper into specific features, and users can explore them in any order.

3. Keeping the list small

With four groups of several steps each, the list could easily become overwhelming. So only one step can be expanded at a time. Users always see the full outline, but only one set of detailed instructions. The trade-off is that they can't compare two steps side by side. Seeing the whole outline mattered more.

4. Help at four levels

Users can learn at the depth they want:

  1. Toolbar tooltips with a short animation and the keyboard shortcut. A simple label appears after 800 ms of hover, then expands into the animated version after another 900 ms, and the pointer can move onto it without closing it.

  2. The task list: suggested steps with progress. Clicking a step expands it and puts an animated, colorful outline on the matching button in the interface, so users know exactly where to click.

  3. "Watch" videos for a full walkthrough of a step, recorded by the learning materials team.

  4. The Navigation & Controls panel: a reference for mouse and keyboard controls.

5. Animations built for handoff

I made 12 animations for onboarding, storyboarded frame by frame in Figma, animated in After Effects and exported as Lottie files. Developers could drop them straight into the product. They scale without pixelation, and small changes can be made in code. That removed most of the back-and-forth that usually happens when animations go from designer to developer.

6. Knowing when to step back

Onboarding that never goes away turns into clutter. I designed the task list to appear in a user's first three blank projects and then stop. Users could close it at any time and turn it back on in Settings.

I also designed progress to be saved per project, not per user, because the steps depend on what exists in that project. Carrying progress across projects would show "Draw a building" as done in a project with no building in it. The cost: an experienced user sees an empty list in a new project. Limiting the list to three projects, and letting users close it, kept that cost small.

What shipped: a simpler version. The panel opens automatically once, the first time a user opens a project. After that, users open it from the home bar. Progress is saved globally and tracked by what the user does while the panel is open, so it carries over between projects.

7. A different entry point for the demo project

The learning team built a demo project: a complex building with many analyses and constraints already set up. On the home page, new users choose between starting their own project and exploring this demo. Showing the task list there made no sense. Every step was already done, so users would see a checklist that told them nothing.

For the demo I designed a simple "what to explore" popup instead, using the familiar "what's new" pattern many tools use for updates. Each highlight pairs a short screen recording or image with a few sentences on what the tool does and what it's good for. I also designed it to be reused for future product updates, so new features get the same lightweight introduction. The learning team agreed with the approach.

What shipped: no separate onboarding for the demo. The PO and the design team decided the sample project already shows everything you can do in Creator, so it didn't need an extra layer on top.

Outcome

Creator launched in early access in October 2026 for Archicad Max and Collaborate subscribers. The onboarding shipped with it:

  • A task list with four groups, starting with the core loop

  • 12 onboarding animations, plus around 20 more across Creator

  • Animated toolbar tooltips with keyboard shortcuts

  • A Navigation & Controls reference panel

Beta testers and internal reviewers responded well to the animations and friendly visual style.

It's too early for usage data. The next step is to measure task completion per group, where users stop, how often people reopen the panel from the home bar, and how often they watch the videos.

What I learned

  • Go back to the original goal before adding control. Locking steps felt safer, but checking it against the product's goal showed it was the wrong call.

  • Plan the ending, not just the start. Deciding when onboarding disappears mattered as much as how it begins.

  • Choose the format with developers in mind from the start. Picking Lottie early saved time for everyone later.

  • Not every decision ships as designed. Two of my decisions were simplified before launch: when the panel appears and how progress is saved. Having the reasoning written down made it clear what each simplification gave up, so the team could make that trade-off knowingly.

  • What I'd change: the "Overall Progress" label actually shows progress within one group. I'd rename it, or show "2 of 4 steps" instead of a percentage, which is clearer at a glance.

← all work

Nemetschek Creator · Early access · 2026

Creator Onboarding

A friendly 3D tool that can do far more than it shows.

What

Onboarding for Nemetschek Creator, a new browser-based tool for early-stage building design.

My part

Sole product designer on the onboarding, end to end: the task list, four levels of help and 12 Lottie animations.

Result

Launched in early access in October 2026 for Archicad Max and Collaborate subscribers.

Role

Sole Product Designer · Onboarding

Team

Product Owner, 3–4 developers, QA, Graphisoft learning materials team

Year

2026

Scope

Onboarding UXMotion designDev handoff

01 — context

A game-like tool for serious work

Creator helps architects, and people with less 3D experience such as real estate professionals, test large building concepts on a real site. Users draw sites and buildings, check area metrics and compliance targets, and run sun, wind and carbon analyses.

Before the product got the green light, I was one of three designers who shaped its early concept. One idea set it apart from tools like Giraffe and Spacio: Creator should feel game-like. Friendly, visual and easy to start. Once it went into production, I owned the onboarding.

02 — the problem

Easy to open, easy to underestimate

The friendly surface hides a lot of depth: drawing tools, metrics panels and several kinds of analysis. The onboarding had to work for two very different groups in the same flow.

Architects

Comfortable in 3D. They don't want to be led by the hand.

Newcomers to 3D

They need to know where to start.

It also had to keep the playful tone. A rigid tutorial would have worked against everything Creator was meant to feel like.

note:

Constraints: the product has its own order (a building needs a site, an analysis needs a building), the content had to fit Graphisoft's wider learning materials, and the animations had to work for developers without me in the loop.

03 — key decisions

A guide, not a tutorial

I started with desk research on how competitors onboard new users. Most use step-by-step tours you have to finish. My first idea was similar: lock later steps until the earlier ones are done. I dropped it, because it went against the product we set out to build.

a

Guide, don't gate

Every step is open from the start. Users can read or watch any step before touching the tool. The only order is the product's own.

locked stepsforced touroptional task list
Locked conceptShipped

The locked concept I set aside → the open task list that shipped.

a

Start with a win

“Create your first project” is already ticked when the list appears. Users start with a small reward instead of an empty checklist.

b

One step open at a time

With four groups, the list could easily overwhelm. Only one step expands at a time, so users always see the full outline. The trade-off: no side-by-side comparison of two steps.

c

Plan the ending

Onboarding that never leaves turns into clutter. I designed the list to appear in a user's first three blank projects, with progress saved per project, and to be closable at any time.

d

A different door for the demo

The learning team built a demo project with every analysis already set up. A checklist there would be all ticks, so I designed a reusable “what to explore” popup instead.

04 — the hook

From a sketch to a wow moment

Together with the learning team, I went through Creator's features and grouped the tasks by how the tools build on each other: Hook, Context and compliance, Unit structure and Optimization. The first group follows the product's core loop; the rest can be explored in any order.

The first group was my proposal. Internally we called it the hook: few steps, all very simple, ending with something impressive.

1

Create your first project

Already done. A quick win.

2

Create your first site

Set the scene.

3

Draw your first building

Simple sketching.

4

Run your first sun analysis

The wow moment.

↑ the hook!

Step 3 in motion: click the corners, close the shape, then pull it up into floors.

Drawing a box is something any tool can do. Watching the sun move across your own building a few minutes later shows what Creator is really for. The list is also built to take new tasks as new features ship.

The payoff of the hook: a sun study in Creator. Source ↗

05 — layered help

As much help as you want, and no more

Users choose how deep they go. Each level only shows up when someone reaches for it.

1

Toolbar tooltips

A label after 800 ms of hover. After another 900 ms it expands into a short animation with the keyboard shortcut.

2

Task list

Suggested steps with progress. Opening a step puts an animated outline on the matching button, so users know exactly where to click.

3

Watch videos

Full walkthroughs of a step, recorded by the learning materials team.

4

Navigation & Controls

A reference panel for mouse and keyboard controls.

task listFollow therecommendedsteps?Explore further steps onthe listStep 1 doneStep 2 active: “Drawyour first site”Hover thehighlightedbutton?Watch the animatedtooltipTaskcomplete?Want moreinfo?Open thedropdown?Details + in-appvideo buttonEnough info?Watch the videoProgress → 2/4Step 3 activeStep 4 pendingAll steps donenext group appearsyesno1/4highlights the buttonyesnoreads the list itemnoyesreveals detailsyesnonoyesnext stepremaining taskscontinueand it starts again with the next groupLEGENDstart / endtask list statewhat the user seesuser decision“no” path

How one task plays out. Each level of help only opens when the user reaches for it.

[Image 4 · add it in the panel]

An animated toolbar tooltip, with the keyboard shortcut.

06 — motion

12 animations, built for handoff

I storyboarded each animation frame by frame in Figma, animated it in After Effects and exported it as Lottie. We picked the format with the developers early. They could drop the files straight into the product, the animations scale without pixelation, and small changes can be made in code. That removed most of the usual back-and-forth.

12

onboarding animations.

~20

more animations across Creator.

07 — what shipped

Not everything shipped as designed

Some of my decisions were simplified or dropped before launch, and I think that's worth showing.

a

When the panel appears

It opens automatically once, the first time a user opens a project. After that, users open it from the home bar.

first 3 blank projectsopens once
b

How progress is saved

Progress is saved globally and tracked while the panel is open, so it carries over between projects.

per projectglobal
c

The demo popup

The PO and the design team decided the demo project already shows what Creator can do, so it shipped without an extra layer.

Because the reasoning behind each decision was written down, the team could see exactly what each simplification gave up, and make that trade-off knowingly.

08 — where it is now

Live in early access

Creator launched in early access in October 2026 with the task list, the animated tooltips, the videos and the Navigation & Controls panel. Beta testers and internal reviewers responded well to the animations and the friendly visual style.

Next to the onboarding, I also designed the project hub, the region selector and the error messages, and updated the full UI kit.

09 — what I learned

What I took from it

Go back to the goal before adding control

Locking steps felt safer. Checking it against what Creator was meant to be showed it was the wrong call.

Plan the ending, not just the start

When onboarding disappears matters as much as how it begins.

Pick the format with developers

Choosing Lottie early saved time for everyone later.

What I'd change

The “Overall Progress” label shows progress within one group. I'd rename it, or show “2 of 4 steps” instead of a percentage.

next: your team?

Right now

Looking for a mid-senior product design role in B2B SaaS or fintech. Available from December 2026.

© 2026 Karola Menyhárt · Budapest

Built in Framer