Making awesome things on the Internet requires the right inspiration. Often, we look to the newest sites and startups for ideas, asking ourselves "what if we made 'startup x' for our industry?" Suppose we tracked, tagged, and detailed these sources of inspiration. Could such a database revolutionize the way we discover ideas? A few months ago, I came across a site from the Biomimicry Institute called AskNature. Some insanely intelligent people have compiled a database of 1400 "strategies" that organisms use to survive against challenges in nature.
The idea is that these "strategies" from nature could solve human challenges in the real-world. Check it:
"Imagine nature's most elegant ideas organized by design and engineering function, so you can enter "filter salt from water" and see how mangroves, penguins, and shorebirds desalinate without fossil fuels."
For the challenge "Protect itself from animals that want to eat it," AskNature has 130 strategies from organisms described in excruciating detail. "Anti-reflective lens" is one such strategy, used by the Hawk-eye Moth as the "inspiring organism."
AskNature is an inspiration database for common problems in science. Instead of trying to re-invent new processes, consider a solar-panel based on a leaf's photosynthesis or build a community modeled after honeybees.
For an engineer, AskNature seems like an awesome resource–hundreds of solutions that are proven to succeed. While I imagine that scientists frequently look to nature for inspiration, the Biomimicry Institute has taken it one extra step by codifying the knowledge.
What's intriguing is that startups, digital strategists–really anyone is business, do this all the time.
How often have you heard, "what if we did something like Pandora?" in a brainstorm. Pandora is AskNature's version of an "inspiring organism" for the strategy "recommendation engine."
"What if we did something like Mint?" Mint is the placeholder for the strategy "visualization" and "aggregation."
"What if we did something like x" is essentially the same mechanism from AskNature. We're asking "what online organisms could we mimic?"
These analogs are helpful; we don't have a term for what Craigslist or Airbnb does in terms of strartegy or function. Going back to biomimicry/AskNature example, "anti-reflective lens" is the strategy employed by the hawk-eye moth to avoid predators. But unless we have the term "anti-reflective lens," phrasing it as "do something like the hawk-eye moth" is a damn good substitute.
"What if we did startup x for client y" is a simple way to do a bit of biomimicry on successful online organisms. It reminds me of Steven Johnson's philosophy on where ideas come from. New ideas are commonly just tweaks on an old idea, but molded into a new context. Without going too deeply on his philosophy, we look to existing startups and websites for inspiration. Just like the coffee houses in the age of enlightenment, a broader exposure to successful startups and websites makes the exercise much easier.
The weird thing is that the most successful online properties and startups aren't necessarily good analogs for inspiration. I don't reference Google, Facebook, ESPN, and Yahoo News to generate ideas. But I do look to smaller companies like Uber, Pinterest, Kickstarter, or Turntable for inspiration. That is, "what if we did a Pinterest for client x?" is far more effective than "what if we did ESPN for client x?"
That is, 'what if we did a Pinterest for client x?' is far more effective than 'what if we did ESPN for client x?'
The difference is that the smaller startups are more interesting in terms of digital mimicry. They are on the edge of innovation, and when I use them as an analog for idea generation, it's far more compelling than the obvious examples. And when a new digital property sprouts up, like Percolate, there's a new to-be-named strategy (perhaps "content idea filtering?") to add to my library of inspiration. The entrepreneur/strategist looking for compelling ideas has an advantage if he or she is constantly searching for interesting sites/startups to mimic, combine, and tweak for a new industry/client.
- Things we care about: strategies from the edge of innovation on the Internet. If the Internet was nature, I care about newly evolved organisms that have adopted interesting/unique strategies to conquer common challenges for survival. That is, referencing a newly discovered microbe that can survive in radioactive waste dumps are far more interesting than inspiration from humans, one of the most carefully studied organisms in history. Note that intuitive strategies such as "walking" and "intelligence" are not entries in the Biomimcry Institute's database.
- Things we don't care about: feature sets. Referencing ESPN's content or Yahoo's brand isn't helpful for idea generation. It needs to be a functional thing, logically fitting into the cliche phrase "it's this for that."
The big question: what does an AskNature database look like for the Internet, a sort of library of digital mimicry knowledge?
Suppose I have a challenge for my users, like "learn new things." What functional strategies does Khan Academy use? And once those strategies are named, we can stop saying, "what if we did a Khan Academy for client X" and start saying "what if we did to-be-named strategy," inspired by Khan Academy?
In short, I'm talking about real-time anthropology/evolutionary biology of the Internet, tagging organisms with their strategies to same detail and breadth of AskNature. Back to Steven Johnson's point: it would be awesome to codify it in one place, especially for the folks who can't stay current with the edge cases of digital innovation.
How hard could it be? Observe how digital organisms are evolving. Tag each organism with a respective strategy. Visualize the data for interesting trends.
The first plan I had was to crowdsource the "it's this for that" relationships for startups. I assumed that it was pretty hard to examine and track digital organisms, so having a big database of connections would make it a hell of a lot easier.
Example: there hundreds of startups that are essentially "Craigslist for industry x." For example, Etsy is Craiglist for arts and crafts. Obviously, Etsy has evolved quite a bit since it started, but I'm sure that at its inception, people compared the site to Craigslist in terms of functionality.
Making a Digital Mimicry Database
I created a little site to try to codify these relationships called StartupDNA.
When you visit the site, you're asked to fill-in an "it's like x for y" statement for a startup. An example: "Think of a site that's kinda like Kickstarter. Then fill-in the statement: ____ is kinda like Kickstarter for ____."
I thought that this would be a clever way to build an evolutionary tree for startups, essentially what the Craigslist diagram above does. Once I had the relationships among startups, I could then begin codifying the functionality that was mimicked. Sadly, it's really hard to complete "it's like x for y" statements. It's not fun unless one has an insane knowledge of startups. The whole mechanic was flawed.
So I dropped the project in favor of a new iteration: WebRecipes.
For several startups, I've replicated the work of the Biomimicry Institute, examining online organisms and attempting to codify strategies that they employ. Pandora, for example, has strategies "music recommendation algorithm" and "streaming music" as a solution to the challenge "discover music," "learn your music preferences," and "personalized radio." To make it generic enough for lateral inspiration, I took it one level higher to "recommendation" and "streaming" as a solution to "discovery" and "personalization." These words became tags. Click on "recommendation" and you can see other startups that employ this same strategy.
The problem with WebRecipes is that it took a shit-load of work. Examining each startup required a good 30 minutes (as it should). Plus, I was really only codifying the startups with which I was already familiar. I wasn't really getting anything out of the database that I didn't already know. I abandoned WebRecipes after adding 30 companies to the database.
So after all that work (and coding), I still have the same vision: just like AskNature, I want a database of digital mimicry, a portfolio of interesting sites that make brainstorming new ideas a hell of a lot easier for clients/startups.
Something to think about: is this doable? What direction haven't I thought of? Is there any value out of a digital mimicry database?
I'm stoked to make something, but I'm not sure if the vision could match reality.
This post originally appeared on Matt Daniels' blog.