Dimo

I wanted to understand why feature-rich, technical software is so hard for new people to learn, and what actually stops them from getting past the basics toward real mastery. Programs like Blender, the free 3D modeling software, are powerful and widely used, but notoriously overwhelming for first-timers. The brief was to pinpoint the specific things that block a beginner from continuing to learn, and then design a better on-ramp than what exists today.

Most drop off within their first few attempts — and it isn't for lack of help: hundreds of thousands of tutorials sit on YouTube, with hundreds more uploaded all the time. So the real question is why, with that much free instruction available, so many beginners still can't learn.

Video tutorials do one part of the job genuinely well: they introduce the software, show off what it can do, and get a curious newcomer excited and started. The breakdown comes after — following along while a video plays isn't the same as understanding, and once it ends most learners can't reproduce the work (Palmiter & Elkerton, 1993; Leahy & Sweller, 2011).

Video opens the door, but it doesn't carry the learner through to mastery.

The overwhelming wall of information gets broken into distinct, structured operations, handed to the learner one at a time. Where a video pours everything out in a single continuous stream, each operation here is something the learner has to actually perform and problem-solve through before moving on. That structure removes the overwhelm and replaces passive watching with active doing.

A web companion for Blender learners who want to understand how to model and not just follow along.

The final product is a working, live prototype that teaches Blender the way people actually retain things: by doing and inspecting at their own pace. Instead of a recording that plays start to finish, each lesson is built from real saved stages of a 3D model the learner can orbit, zoom, and study from any angle. The model's structure stays visible the whole time, which is exactly what a video can never pause long enough to show.

Community research

Read through popular Blender threads where learners openly vent about being stuck in "tutorial hell." The sentiment was strikingly consistent: people feel overwhelmed and discouraged, unsure why they still can't make anything on their own after hours of tutorials. Hearing that frustration first-hand made the problem feel real, not just academic.

Secondary research

Dug into 30 years of research on how people learn software to understand why video tutorials leave learners stuck. The number that stuck with me: across 6.9 million viewing sessions, people tune out of instructional videos after about six minutes no matter how long they are, and most Blender tutorials run 10 to 40 minutes (Guo, Kim & Rubin, 2014). Other studies showed the same pattern — people can follow a video in the moment but can't redo the work a week later (Palmiter & Elkerton, 1993; Leahy & Sweller, 2011), and the hardest part isn't any single video, it's finding the right next one and applying it to your own project (Drosos, Sarkar & Gordon, 2024).

Market research

Looked at how other teaching platforms succeed by having people learn through active, hands-on practice instead of passive watching, with Kaggle as the clearest example. It showed that a do-it-yourself approach already works in other fields, and pointed to what a better Blender learning experience could borrow.

The core bet paid off. Watching a first-time user rebuild a model unaided after one pass was the clearest signal that active, inspectable learning does what video can't. Keeping the model's true structure visible turned out to be the heart of the product, the one thing a learner can't get from a screen recording.

What could be better

Testing surfaced real gaps. Beginners needed clearer step-by-step guidance tied to each shortcut, plus help simply knowing which mode they were in, something Blender itself does poorly. A few early ideas, like smoothly animating between stages, had to be set aside because they fought against keeping the structure accurate.