Simplicity Rules

Do humans return better search results than computers? Yes, actual intelligence beats the artificial kind. But computers are really good at aggregating what humans have created.

There’s talk about social search being the next wave. Google already uses social search. They call it PageRank, and it’s why they are number one right now. It changed the game, and aggregated our opinions about sites in a usable way.

Social search may be the past, but it is also the future, as long as computers are still involved:

1. With the web opening up, human-filtered content will become more available to search engines. Check out the one result search query for an example.
2. Personalization is another method to filter what humans are already creating. See Greg Linden’s recent post characterizing value of personalized search. Linden wrote the first Amazon recommendation system.
3. Leveraging what my friends like might be the purest form of social search. For this to work, it’s going to take more than me sharing a single link. Sites I have liked need to help my friend without my immediate intervention. While social networks might appear best poised, it’s going to take data, so someone like del.icio.us is actually closer than Facebook, for example.

As much as some might despise game-able algorithms determining search results, that’s not going away. Even with social search, we still need the computers to do the heavy lifting.

It’s a common misconception that anyone who works on the web is a graphic artist. I haven’t been a web designer since back when that’s all you could be as an HTML monkey.

I make the web work. Getting to that point, I’ve tried several titles: Webmaster, Web Developer, and finally Web Programmer. And still many people, even those who should know better, think I’m a designer.

And the thing is, I play one. I’m really lucky to have worked with talented designers. Their work looks great, and is also intuitive and usable. But even if I have one of these superstars beside me in the trenches, there will always be design decisions I need to make on my own.

1. I added a tiny new piece of information to a page–where should it go?
2. We’re changing ad sizes–how will that affect layout?
3. Nobody seems to know where to sign up–should we call it “registration” instead, or change the location of the link?

While I continue to refuse the label of designer, I am also immersed in the high-level topics a designer cares about. We cannot completely separate the responsibilities of programmers and designers. They need to work together and even know each others’ turf.

How a user expects a product to work is as much about design as it is about code.

Recently I wrote about the difficulty choosing a camera among many choices. Essentially, the problem was that Canon had several cameras that were essentially the same. In the comments to my original post, Jason gave a reason that manufacturers might have so many models:

“I’m sure GE would rather produce one product, but the large stores refuse to carry products that aren’t unique, in order to give them pricing freedoms, and not be held to their price matching promises if a competitor runs a special.”

Given that they must create so many similar products, it makes sense that companies resort to pseudo-random product numbering. The side effect is that not only can I not decide which item to buy, once I do, there’s no way for me to understand its place in the world. If we all went by our social security numbers, I might not have as many friends that I care about.

Des Traynor says to give your products names, not numbers, using mobile phones as the largest example:

“Funnily enough, the only people I know who can say it with certainty own the Motorola Razr. It’s easy to write that off as a result of impressive advertising, and a distinctive look and feel. There is one other big reason though, the Razr has a name.”

There are some humorous counter-examples in the comments (Levi’s, BMW, Chanel perfumes), but the point remains: names mean something, numbers don’t.

Here’s an idea that will help you save everybody’s time by applying constraints. Crunching your own time is easy. Just have a POWER HOUR, Four Day Work Week, or create a Seven Day Product.

To save others time, try time-boxing:

Time-boxing is “the setting of artificial time constraints for tasks like brainstorming and issue resolution. The objective is to cut down on exhaustive consideration of endless possibilities. With pressure to stay focused and disciplined, we can reduce the amount of time it takes to reach consensus.”

This idea comes from Ryan Freitas in an article comparing designers and chefs. For all four kitchen analogies, read the three page article Cooking Lessons for Designers (PDF 850K).

The real new web lies in off-the-cuff self-expression. Some are calling this microblogging.

It is easy for anyone to participate, because it easily integrates into your current activities:

• If you find a link on the web, share it (with friends or the world). Possibly include a bit of commentary about what you think. Example: del.icio.us
• Let people know what you’re up to. Answer from the web or your mobile phone. Example: Twitter, Facebook status.
• Upload a picture to a service that resizes and shares it. Easier than emailing digipics. Example: Flickr

And there are probably many other examples, with lots more added all the time. Each will be a tiny piece of letting me express myself during my regular routine. If it takes too much additional effort, 1) I might not do it and 2) it’s not really off-the-cuff.

The missing piece, which some are undoubtedly working on, is the connector between these, something to aggregate your digital life. Someone will figure it out and then there’ll be better ways to get a complete view of someone in near real-time.

Once these rivers of self-expression are flowing, the death of permanence will be even more obvious. Which leads me to wonder… as much as I like this off-the-cuff stuff, wouldn’t it be great if more permanent expression was just as easy?

The numbers out of Facebook are staggeringly large:

• 0ver 40 billion page views in May 2007
• Over 150,000 registrants daily.
• #1 photo sharing app on the web. 2.7 billion photos on site.

Many would point to these and say they are an example of amazing scaling. Instead, another jumped out at me:

• Half user are outside college. That number was zero in Sept. 2006.

Facebook started in 2004 at Harvard. When they got that down, they added three more universities. After over a year, the site expanded to cover 800 schools. Then they added high schools, they went international, and added work networks.

Finally, after they had things down pretty well, and after two years and eight months, Facebook opened up for everyone to join.

It’s not only about server scaling. It’s about idea scaling.

Do you have big ideas and dreams? I do.

I also believe that ideas alone aren’t worth much. Execution matters.

From there, you might get that the game plan for success is to jump completely into a huge undertaking. Unfortunately, you could end up biting off more than you can chew.

Tenet to live by: be wary when there’s an idiom about your situation–you may end up screwed.

I have done plenty of jumping in. I’ve been burnt and I’ve become burnt out. As great as big dreams are, you need to take small, simple steps to get to them. You need to worry about scaling. Yourself, your servers, but most importantly, your concept.

Whatever your big idea, other people don’t get it all at once. You might not even really get it until you start working on it.

In Mathematics, there’s a type of proof that uses induction. You start with a base case and show that something works for it. Then, you show that if something works any step, it will work for the next step.

Put that to work on your idea. Get something out, because that’s vastly better than nothing. Then see what you can do with that something, and how the next thing fits.

When something costs nothing, you’ll get a lot more people to try it than if you charge up front. That’s the idea behind freebies, on the web and otherwise. Free lowers a barrier to entry.

Consider this:

1. Shopify had a service that allowed anyone to create an online store for nothing up front. Shopify only got paid when the seller was paid.
2. Then they up and added a monthly fee.

Portland Web Innovators has a discussion of the changes. Poor communication from Shopify seems to be the biggest culprit here, but the long-term damage may come in the giant barrier they constructed in front of prospective customers.

I think we’ll see a free version back soon, with premium features available for a monthly fee. The “Freemium Business Model” is powerful because it harnesses the simplicity of free and adds on important bits for those who get the most benefit from your product.

The model allows new customers to hold off on paying until they know it’s worth it. Money now may be better than money later, but any time you ask for money, a big chunk of folks leave.

A Portland company that plays the freemium game well is SurveyMonkey. Their price list shows a good free service with room to move up.

Throughout their site SurveyMonkey reminds the free users of features they could have. It’s like a shareware registration nag, but less intrusive.

Those who are in it for the free stuff keep on taking, but who will they go to when they need more than 100 respondents or want to redirect to their own site upon completion? Most of them will choose SurveyMonkey, because they’re already familiar with what it can do.

Shopify may go for the harder business model. Maybe it will do better, but I doubt it. Free is very easy, and on the web simplicity–and its low barrier–rules.

I’ve been saying this for a long time. Maybe even back before blogs. So, when Tom made his Case for the Crusty Old Website, I couldn’t help but cheer.

“I’m not really trying to be a curmudgeon and reject the new outright, but last week I ran across Chris Glass’ website it reminded me of how much fun I had exploring peoples personal sites just a few years ago. With all the hype surrounding new social networks, a case can and should be made for those, you know, personal websites, that we used to visit. Blogging and the internet is, among many other things, just one big social network.”

Even though I really like the web as it is today, I can’t help but be nostalgic. Personally, I miss putting together little proofs of concepts or silly play-things. And when those sorts of projects do happen nowadays, people tend to give them their own vanity URL, because even the smallest mashups might become something bigger.

Like Tom, I miss exploring the partially organized bits that people have placed online. A blog feels like a stream of content flowing away from me. No matter how many permalinks you have, it won’t be as permanent as the little bits we put together without a Content Management System.

But, well, now I’m feeling like a curmudgeon, too. Like I said, I can’t help but be nostalgic.

No matter what anyone does, says, or thinks, Barry Bonds is one swing away from the most prestigious record in baseball. I’m pretty jazzed about that, not as a Giants fan (though I am one), but as a baseball fan. I’ve enjoyed watching Yahoo!’s Major League Baseball page track the homers.

I’m also a web geek, so I peeked in at the code and realized I could see into the future.

I looked at the URL of the image with Bonds in his current state, tied with Hank Aaron. As a test, I changed out a couple of characters and was surprised to see Yahoo! already has the record-breaking images on their server.

So, whenever Bonds breaks the record, we’ll see one of two options on Yahoo!’s MLB site:

Personally, I hope it’s the fireworks graphic. Regardless of whether you feel Bonds’ record is rightfully tainted, the feat should be celebrated. It’s been 33 years since Aaron made his mark, and it will be at least seven years before we see it broken again.