Here’s the first image I sent him, along with the “final” version.

I realized, for the first time, that my initial sketch was absolutely terrible. I’m surprised he ever said okay to it.

Back when I first started the painting, I saw the guy on the right when I looked at the guy on the left. Sure, there was no colors and it was all rough, but it was there. The reflection in the goggles. The stern look. The lingering, unexplained story of the destroyed starship in the goggles. Even now when I look at the guy on the right, I’m probably seeing an even better version of what’s there.

I had the same affliction, as a digital painter, that a lot of startups have running their business: Vision blindness. It’s when your own vision, whether it’s for a business or a digital painting, blinds you to reality.

This oxymoron of a term is crucial to startup success while also being a startup’s greatest liability.

Vision blindness: The prognosis

Unfortunately, most entrepreneurs can’t see through their own vision when looking at their product. I’m certainly one of them, and there’s no shame in it. If you’re reading this, you’re probably vision blind too. (If you aren’t, then you probably aren’t passionate enough for startups. Go find a job at a bank.)

Google definitely had vision blindness when the rest of the world said “We don’t need another search engine.” In a way, it can be an asset. Most successful startups have been told they will fail. Most successful startups have had a crappy product at one point.

And most successful startups have vision blindness. Even after their success.

If you take the first day of Google and put it next to today’s product, the difference would probably be as stark as the difference in my two paintings. Of course, a startup has more qualities than a painting– vision blindness doesn’t just concern the visual.

Treatment #1: Iteration

There’s a month’s worth of spare time in iterations between my iniital piece and my finished painting. Quite a few hours of hard, focused work.

A lot can be said for just making a product better and better. Don’t confuse this too much with adding to a product. Adding things can make it better, sure, but they can also make it worse. When people talk about taste, this is what they’re talking about: Correctly iterating.

Iterate, iterate, iterate. You’re a startup. It costs you basically nothing to iterate, so get to work.

Treatment #2: Feedback

Feedback is one the hardest things to find in startups. Your friends and family will endlessly lie to you. My dad has been completely fine with every iteration leading up to the last one. Most of them sucked. Badly.

Don’t fret, though, they do it because they love you. Or because you’d raise a big stink if they were honest (You don’t do  that, do you?).

Dishonest (even it’s polite, white lies) feedback will make your vision blindness worse. Avoid it at all costs.

Instead, ask people who hate you. Let people be anonymous in their feedback. Ask someone who has no stake in you or your product. Do whatever you can do get good feedback. When you do, assess it as honestly as possible; remember that you’re blind and they aren’t.

Feedback is one of the best reasons to have a co-founder. If your co-founder gives you polite, easy-to-give, white-lies feedback, then find a new co-founder. Or stop being a dick about receiving the feedback. (And yes, the problem is either him or you.)

Treatment #3: Shifting Perspective

You can temporarily fix vision blindness by shifting your perspective.

A trick in digital painting is to flip the image horizontally. The image is new to your mind, so you see a lot of glaring errors that weren’t there before.

Unfortunately there isn’t a great analog for startups. One way is to try to imagine how your product will be used by all the different types of people who use it. Think about the system (your product) not as an overall entity, but as something approached by a single person using it.

Watch someone use it, too. Flaws and errors show themselves really fast that way.

Treatment #4: Abandonment

Cures for vision blindness are pretty rare. The sure thing is abandoning the project. If you drop your vision, there’s no way it can blind you.

“But what if I’m right?” is the question that comes up here. Your Magical Fart Facebook app might be the next Google, sure. Or it might be a huge waste of time.

For every naysayer who says your idea won’t work, there’s an even stupider idea that made someone a millionaire. There’s no way to really know if your idea is good. Stick with it for a while and feel it out. If you’re treating your vision blindness, you should have a pretty good idea pretty fast.

Of course, this is coming from a guy who’s built products for social media newsrooms, educational tech, product placement, and more and only given up on one of them. Ultimately, we’re the own keepers of our vision.

In the end, though, if it’s not working, drop it. If it is, keep it up. There’s no shame in failure. Just consider it an iteration in the overall picture of your startup success.

Concept art for the three basic classes of the first Diablo.

The Warrior

As the heavy hitter (he’s the guy with the big ax in the picture), the warrior ends up doing a lot of the damage.

But he’s more useful for the damage that he takes. A team of adventurers will have a hard time functioning without a crazy guy up front, swinging around his weapon and taking all of the big hits.

In web development, the warrior is the back-end programmer. It’s a dirty, bloody, sweaty job, but somebody’s gotta do it.

The Mage

After slinging around a few magic missiles, the mage is the guy on the team that gets to act like the whole thing was no big deal. He drops a few tidbits of wisdom on awed onlookers in his fancy robes while the warrior gets to pick bits of ogre out of his ragged warrior beard.

That’s right; the mage is the designer. Everyone see the lights of his fireballs from the village, and comes running to congratulate him on a job well done after the village is saved.

But if a Goblin were to slip by with a spiked club, the mage and his fancy robes wouldn’t last long. The Mage has the hardest time on his own.

The Rogue

The Rogue is the most versatile of the three classes. He can attack from afar, like the mage, with his bow and arrow. He can also mix it up in the front lines with his dagger, particularly if he can position himself correctly (ideally, behind his opponent).

It’s this versatility that lends the rogue his real strength. He can be where the team needs him to be.  In web development, the rogue is the guy with an unusual combination of skills. He’s sometimes called a sweeper or just a jack-of-all trades.

Building your team

Putting together a web development team is just like putting together a band of adventurers. You need a balance of classes.

If you have three people, aim for one of each class. Sometimes it’s okay to have two warriors and a mage, or similar combinations. Having three of the same class is almost always a disaster.

With two people, the combinations become a bit more limiting. Obviously Warrior/Mage or Warrior/Rogue is a good combo. Notice that having at least one warrior is always a good idea (the converse, having all warriors, is just as bad as that idea is good, however).

Watching out for the know-it-alls

Adventurers are a boastful bunch. Watch out for the warrior who says he knows everything there is to know about magic, or the mage carrying around a great sword that’s taller than he is. For every day an adventurer spends learning more about the intricacies of shooting fireballs, the less he learns about swinging an ax. There’s lots to know in fantasy adventuring, and there’s even more to know in web development.

We’re long past the days of one person being able to know everything. A balanced team of specialists will always beat out a team of know-it-alls.

It’s all about taste.

Ramen noodles are the lifeblood of my small startup. Their ease of preparation and low, low price make them ideal for bootstrapping. I had perfected my cooking technique and couldn’t be happier with my ramen noodles.

Or so I thought!

With the iPan, Apple have once again introduced a complete game changer. Preparing and eating ramen noodles will never be the same.

Apple’s saucepan brings something to the table that other saucepans just can’t match: Taste. iPan ramen just tastes better. The taste continuum below perfectly represents why Apple products, especially the iPan, are superior.

Not just a better development environment: A complete development environment.

Before, my development shop was plagued with the worst kind of problem: Too many choices.

Now Apple have done even more work to lift that burden from my shoulders. I no longer have to go to the store and pick from several different saucepans, with their erratic problems and glitches. I have the iPan.

In fact, I couldn’t go to the store and pick if I wanted to. Apple have extended the terms of their agreement to include the complete development environment– which now includes our kitchen.

Some companies would be content to just provide an alternative. Apple thankfully mandates our development environment, ensuring a great experience for everyone.

They go that extra distance so I don’t have to.

Apple have also quite cleverly ensured that I will always pick ramen noodles. After dropping 1700 on a MacBook, 500 on an iPhone, 700 on an iPad, 99 a year to develop for the iPhone, and 149 on this revolutionary new iPan, all I can afford is ramen.

Thanks, Apple.

I’ll repeat that: Don’t be evil.

I stress this because the motto is one of those phrases that gets exaggerated given some time to circulate in the community. Most frequently it’s twisted into “Do no evil.” It’s like Kanye’s famous quote that he never said, “George Bush hates black people,” and Darth Vader’s ever-popular unquote, “Luke, I am your father.”

In Google’s recent purchase of Etherpad, a lot of people got nervous. They called foul since Google was violating their own nonexistent corporate policy of “Do no evil.” They weren’t. One can do evil and still not be evil, given the right mixtures. If Gandhi had kicked a kitten at some point in his life, I’m sure we’d all still consider him an alright guy.

Google themselves have commented on this, back in 2006 when they tangled with Chinese censorship. Schmidt said they put together an evil scale to make sure they were still in the clear, and that not serving China at all would be a greater evil.

Now it’s a moot point, of course, since Google has decided to open-source Etherpad instead of completely crushing the project.

But still, let’s remember to hold Google to the standards they actually espouse, and not the ones that we make up.