There is a specific disappointment you feel the first time you open a wiki, land on the "how to contribute" page, and realize you have to read 5,000 words of policy before you can fix a typo. This is not a bug in the wiki. It is the natural result of lowering the barrier to entry without addressing the follow-up question: okay, but what do I actually do?
The open source world has a standard answer to this problem. Lower the barrier further. Make it easier to edit. Add a WYSIWYG editor. Reduce friction. The assumption is that contribution is a gradient problem: if we make it 5% easier to contribute, 5% more people will contribute. And this is true, up to a point. But it misses something fundamental.
I live in Berlin and I have walked past the same bench on my corner about four hundred times. I have never once wondered whether it was made of wood or metal, whether it had armrests, whether the surface was in good condition. I sat on it. That was the extent of my relationship with that bench. Then I installed StreetComplete and the app asked me to answer exactly those questions about exactly that bench. I walked over, checked, tapped my answers, and watched the bench on the map change from gray to a satisfyingly colored icon.
This is the coloring book principle. A blank piece of paper is paralyzing. A coloring book with numbers on each section is not. The numbers don't make the task possible—coloring between the lines was always possible. They make the task obvious. They tell you exactly where the next action is and what it should look like.
Wikipedia's empty articles sit on the internet for years. GitHub's "good first issue" labels try to point beginners toward a task, but they require you to understand the project's domain, read the issue description, and decide if you can handle it. The label doesn't point. It categorizes. There is a difference between telling someone "this is a small task" and telling them "click here, type this, done."
StreetComplete does something more radical than either of these approaches. It doesn't just lower the barrier. It eliminates the question entirely. The app shows you a map of your neighborhood with numbers on every unpainted building. The numbers are quests. You walk to one, answer a question about it, and the number disappears. There is no moment where you ask "what should I work on?" The app has already answered that. It has distributed the work across the geography of your daily life, and it has made each piece of work so specific that only you could answer it.
The locality multiplier is the hidden engine here. StreetComplete gives you tasks that only you can answer, because you are standing on that street, looking at that shop, checking that bench. Wikipedia's suggested edits give you generic tasks on topics you don't know. Being the only person who can answer something is a stronger motivator than being asked to help. It makes you feel like a witness rather than a worker.
StreetComplete also rejects the entire gamification playbook. No points. No badges. No leaderboards. No streaks. The only reward is the map changing color. A gray street turns green when you have answered all its questions. This sounds like it would not work. It works frighteningly well because the map changing color is a direct representation of your effect on the world, not a tokenized abstraction of it. Points are a proxy. Green is the territory.
The obvious critique is that this approach has limits. StreetComplete's quests are finite. Once every building on your street is mapped, the app has nothing left to ask you. The coloring book fills up. Wikipedia's empty articles are inexhaustible because knowledge is fractal. StreetComplete's quests are bounded by the physical world. This is a structural constraint on the model: the system organizes contributions so efficiently that it exhausts its own territory. Success is its own completion condition.
There is a darker edge too. StreetComplete centralizes the question of what matters into the hands of quest designers. Subjective knowledge—is this bench comfortable? does this street feel safe at night?—gets excluded not because it is unimportant but because it cannot be structured into a multiple-choice question. The filter is not value. It is verifiability. The coloring book model can only ask questions that have predetermined answers.
And there is a trap in the model itself. StreetComplete gets you to contribute a million house numbers, but OpenStreetMap still needs someone to trace the aerial imagery of the new housing development. The model that gets everyone to contribute small things also keeps everyone's contributions small. You don't graduate from quests to real editing because the quests are the whole point. The onramp becomes the destination.
I think the coloring book principle applies far beyond mapping. Language learning apps like Duolingo structure a vast, intimidating skill into discrete, ordered lessons. The daily prompt in journaling apps tells you exactly what to write about. Even something as simple as a README with a "quick start" section follows the same logic: don't explain the architecture first, tell me exactly which three commands to run and what I should see afterward.
The lesson is not about lowering barriers. It is about structuring the void. A blank canvas is not freeing. It is overwhelming. The best contribution systems do not say "anyone can help." They say "walk to the bench on your corner, look at it, and tell me what you see."