The AE Principle

The first paragraph of any paper should be the easiest to read. Registering for a service should be fast. Distribution should start locally. Games should have intuitive controls. A professor should be casual on the first day of class. etc.

Why? Because people like it when something requires a low Activation Energy (AE).
.

What is AE?

.
There are three prerequisites for a chemical reaction to occur. For one, the molecules involved must physically collide with one another. Second, those molecules must be oriented properly at the time of collision. Finally, there must be enough energy in the system for the two to collide. This third point refers to a reaction’s “activation energy.”

The basic premise is that it takes an initial shock to get a reaction going. If this weren’t the case, shit might be exploding around us all the time. Luckily it takes things like catalysts and enzymes to lower the activation energy enough to the point where useful reactions, like the ones that maintain healthy levels of oxygen, can occur consistently.
.

Activation Energy from Wikipedia

.
I’m not really interested in chemistry, though, and I don’t know enough about biology to discuss it with confidence. But I think we can use the principle of AE to address other important things, like addiction, group dynamics, voting, and even how Internet Explorer is more popular than Firefox despite being a noticeably worse product [1].

For the most part my analysis will seem intuitive, even obvious, but that’s part of the point I’m trying to make. People understand AE and often take it for granted; I’m going to argue that thinking about it explicitly might help entrepreneurs get their products off the ground, or inform activists about voter turnout, or teach professors how to retain kids in their classes. Indeed, AE - a vital component of the chemical reactions underlying all life - is a fundamental variable in all kinds of familiar systems, and one which deserves closer attention.
.

Some examples

.

‘That’ professor

.
Anybody who’s been to college has had ‘that’ professor, the one who challenges them, motivates them, and ultimately, teaches them more than the syllabus ever said they would. I think all these professors share three traits in common: they’re demanding (they expect good work, give As sparingly, assign a lot of reading, and ask difficult questions), rigorous (they thoroughly address the material and go “behind the scenes,” so to speak), and funny.

It’s important that professors establish these things early on, often within the first week of class. At many colleges this first week is designed purposefully for “auditing,” or previewing, classes. Students, especially as freshmen, are often actively encouraged to sample a wide variety of classes across disciplines. Even if this isn’t an explicit concern, most colleges have a drop/add deadline which gives students a few weeks to try a class, drop it, and replace it without penalty. The last thing a professor wants is to be the guy whose class is empty by the time that auditing period is over; not only would it be embarrassing, but often research dollars are tied to class performance, so it could hurt their bottom line as well.

There are a couple of ways good professors ensure this won’t happen. For one, I’ve noticed that they swear. By just casually saying “fuck” or “shit” during the first lecture, they put themselves on par with their students. Kids respond better to a person than a professor; they want to hear anecdotes, stories, swear-words and jokes more than bullet points and topic sentences. Dropping the F-bomb is one of the easiest ways to lower the AE of listening, which is the biggest impediment to the true challenge - namely, learning shit.

They also show, within the first hour or so, that they’re going to be demanding and rigorous. Their syllabuses illustrate how heavy the workload is going to be, but they avoid dwelling on procedural details that students dislike most about formalized education. They talk about the textbook, and even bring a copy in for students to peruse so they aren’t forced to pay $150 to find out they hate the material. They get started on “actual” work in the first class, and even distribute a problem set or writing prompt, to show students what they can expect. In short, they do things that lower the AE of investing in a class; they make the course opt-in rather than opt-out, laying everything out flatly and concretely and inviting students to join them. When that happens, attending class is less about figuring out what it’s going to be like and more about jumping right into the topic of interest.

Interestingly, I’ve found that a course’s AE is often inversely related to its overall energy requirement. In other words, a high AE often indicates a slow and easy class requiring little effort, with the opposite holding for courses with low AE. The key point for me is that in a lot of cases I almost definitely wouldn’t have invested so much in a professor, course, or topic had it not been for a comfortable start. So a low AE got me, and could get most students, to do things they didn’t know they could do, since it’s often the initial push that’s the hardest, just like the hardest part of running is getting up and putting your shoes on.
.

Addiction

.
Lowering AE isn’t always a good thing, though. For instance, I now smoke a few cigarettes a day because it was so easy to start. Although there is such a mountain of information that discourages smoking, it’s almost nothing compared to the simple “try one” from a friend. Indeed, most of that mountain is about long-term effects and aggregated damage; the “try one” is a quick hit, a short-lived experiment that’s easy to do right now.

The problem is only compounded by other social forces. While smoking used to be mainstream it has now developed into a kind of counter-cultural thing, especially among kids, for whom most of the anti-smoking campaigns are designed. Smokers stick together, and pulling one out at a party almost guarantees a good conversation. People working against a common adversary (the non-smoker), especially when they’re in a minority, have an easy time getting along. Indeed, as weird as it may sound, I met many of my good friends outside a building with clouds of toxins hanging around us.

The AE principle, then, has solid foundations in addiction. It certainly explains some phenomena. For instance, while heroin is obviously more addictive than smoking cigarettes or marijuana, it has a relatively extreme AE: not only is it poorly distributed, it’s downright scary; people are a lot more afraid of getting hooked and dying in a few months than getting hooked and dying in forty years or, in the case of weed, becoming lazy and wanting to eat a lot.

Thus campaigns against addiction should seek to raise the AE. Oftentimes they do this: making cigarettes expensive, selling them behind the counter, and restricting them to a segment of the population are all effective measures. In other ways simple rhetoric has helped - killing the cool factor of cigarettes discourages experimentation, and propounding the dangers of smoking increases that initial fear level. However, many of these initiatives only increase smoking’s “counter-cultural” appeal and pit the smoking community against an ever-growing adversary. Addressing these collateral effects should do a lot to help; an example might be a campaign against “social smoking” or simple measures that make it prohibitively expensive to share cigarettes or severely limit their distribution. These seem less promising, though, when set against a powerful engine (the tobacco industry and the government that accommodates them) seeking to maintain the AE as it dwindles.
.

Facebook

.
Addiction isn’t necessarily destructive. Indeed, many businesses thrive because they’ve successfully captured and addicted a large number of people. Their key realization, conscious or not, is that a low AE goes hand in hand with, and often encourages, addictive behavior among their users.

Take Facebook. At this point their user base is in the tens of millions, and I’d argue that at least half of them are in some sense addicted to the service. I have friends who check it more than 15 times a day; when I go to my campus’s computing centers, nearly a third of the people are using it actively or in the background; kids increasingly “can’t live without it” when it comes to sharing gossip, pictures, and news of events or parties.

How did Zuckerberg (the young founder) et. al. get there?

Facebook’s AE curve is notable in that it is inversely correlated with its user base. In other words, as more people use Facebook, it becomes easier for others to join - once your friends are using it, you’re practically *forced* to sign up, as the all-important legwork of gathering people with whom you’ll trade information and messages is already done. The inherent caveat is that the user base we’re talking about is only local - people don’t care about the millions of strangers on the service; they only care about their friends or acquaintances from school. So Zuckerberg didn’t have to attract millions of people at once to create value; he could do it one group at a time, which is in fact much easier.

Because Facebook is nothing more than a disjointed network of networks, Zuckerberg just started with his own at Harvard. The network quickly built itself as friends invited others to join. The “invitation” feature was, at least in the early days, Facebook’s most valuable. It encouraged real community; the easiest way to join Facebook was to be invited, and that meant you already knew at least one person in the network. Since your friend was already in (this helps you trust the service) and you only had to click one link in an e-mail to get started, you had essentially no AE. The hardest part of registering was coming up with stuff to put on your profile; this turns out to be not much of an impediment since people (including me) love talking about themselves. The result was that instead of being a place for a bunch of strangers to get together and talk about some thing, like most social websites at the time, Facebook was comprised of a bunch of closely related people gathering for no reason at all and to discuss nothing in particular.

As users joined in this way, and as the AE for new users dropped, something else happened - the service became concomitantly more addicting to the people already using it, since now more of their friends got involved, and more content (including messages, pictures, wall posts, interests, files, and blog posts) was created and distributed. Then as these users spent more time on the site, they increasingly spread it among their friends (or their friends caught them checking it frequently and said “what’s this?”), and the cycle sort of repeated itself.

At this point, at least among college users (Facebook has since expanded), Facebook’s AE curve has leveled out. The network is essentially built out, and anybody that would have joined has joined already. But be sure it’s going to be around for a while, as new classes of high school students catch on to the phenomenon and get addicted themselves. The AE is so low now it’s bound to happen (during their first week of school people will be asking them for their full names so they can find them on Facebook).
.

YouTube

.
Facebook is by no means unique. Community-building is at the heart of almost every web-based business today, and they’re calling it “Web 2.0“. The driving force behind this movement are the tools we’ve already talked about - invitations, sharing, etc., and YouTube is a great example of how to use them properly.

YouTube was built to make web-based video work. It took one great innovation and used the same tools as Facebook to spread it, reinforce it, and turn it into the monolithic King of New Media it is today.

The guys who started YouTube realized that when it comes to video, quality doesn’t matter. People don’t want pretty, they want it now. Technically, this means avoiding proprietary formats like Quicktime (.avi) and Windows Media (.wmv) that some people don’t have and opting for the lower-quality, but faster-loading and ubiquitous Flash video format (.swf). The challenge was allowing users to upload videos in any format, even from their cell phones, and convert it quickly into .swf’s for distribution. Once they figured that out, it was off to the races.

YouTube built in the link-sharing, HTML embedding (the code snippets they allow you to copy and paste into your website), commenting, “e-mail this” buttons after the movie runs, etc., to let people take those videos and spread them across the web. But unlike Facebook, who uses all those same tools, YouTube users are anonymous, independent, and generally uninterested in the site itself or community around it - all they care about is the handful of videos their friends send them or that they saw while surfing the web. But since the AE of sending a few YouTube videos to your friends or even posting your own is so low, millions of people do it all the time. They might not spend as much time on the site as on Facebook, or be as “into it” as Facebook users, but when you multiply millions of people by a few pageviews, you get something worth about $1.6B.

Low AE seems to be pretty important component of a lot of successful endeavors, from teaching a class to building a multi-billion-dollar web presence. For each of these success stories it’d be trivial to find ten more that failed because they ignored their users’/customers’/students’ inherent unwillingness to get started. Unfortunately most of us have never heard of these failures. But enough of that. How can we use the AE principle?
.

Applying AE

.
I might jump around a little, but I want to try to show a broad set of applications here, rather than focusing too much on one type of problem.
.

Zeedex

.
I subscribe to a newsgroup for people who attended Startup School in Cambridge, Mass., back in ‘05. It’s basically a collection of really smart people discussing ideas for young companies, looking for partners or co-founders, and seeking advice from the industry gurus who occasionally drop in and offer their input. Recently a kid came on there and asked for advice regarding his new website, Zeedex.com. Zeedex is a place for suggested search listings, where you type in some keyword, like “computer” or “dog,” and it returns a list of related keywords along with traditional Google results. Anybody can edit these lists, much like anyone can edit a Wikipedia article.

The main problem with the site is that, since it’s so new, nobody really uses it, and so for most search terms there are no corresponding “lists”. There are simply not enough users to do that, and one kid and his friends can only add so much. So Russell’s (the founder) main concern was gaining penetration, or making his service popular enough that it would add value to the typical person’s search. How could he do it?

Most of my response revolved around AE. It was important, I told him, to lower the AE of adding a list so users would be encouraged to do it. As it stands a user has to actually do a search for something to edit its corresponding list. Since there are literally an infinite number of possible search terms, this doesn’t seem that practical. So to improve the situation, I suggested he tap into an idea behind Wikipedia that most people who read it don’t know about.

Wikipedia has what’s called a Community Portal, where it houses its Talk, Project, Featured Article, and other pages. The point of this portal is to (a) give people concrete tasks, in an easily accessible place, for improving the Wiki. For Zeedex this might mean an area for highlighting certain popular search terms that need new or better lists. (b) It gives users concrete goals, and a chance at notoriety. People are proud of their tidbits of knowledge and most appreciate the recognition that comes from sharing them with the world. So in the Community Portal a bunch of people might rally behind a “Trigonometry” page to get it recognized as a Wikipedia “Featured Article”, and more often than not, people take this seriously enough to devote a lot of time to the project, and the Wiki improves as a result. The people who contributed are often recognized formally (with things called “barnstars”, no joke) or informally through comments on one of the many Talk pages. For Zeedex it might be useful for people to see how many searchers clicked their link, or how they rank among other submitters, for instance, so they get a sense of their contribution’s impact. In effect these initiatives simply give people incentives to get involved, which helps break the biggest AE impediment: “Why would I waste my time doing x?”
.

Video conferencing

.
I find it interesting that something as cool as video conferencing, that’s been around for so long, hasn’t really taken off for your average computer user. A couple of the earliest AE bumps, like bandwidth restrictions and a lack of software support, have been taken care of for years. Indeed, most chat clients, including AIM, iChat, ICQ, MSN, Skype, etc., have video conferencing functionality built-in, and a good proportion (around 50%) of Internet users have broadband connections. What’s the problem?

Nobody has a camera. The thing is, it makes no sense for most people to go out and buy one, even if they’re cheap (~$50), if their friends don’t have them. The AE is way too high. Recently, though, computer manufacturers including Apple and Dell have made moves that lead me to believe videoconferencing will gain new life in the next five years. The latest line of laptops from both companies (the higher-end ones with Dell) now include cameras out-of-the-box, built into the space above the display. Once we go through a few computer life cycles, which are on average 2-3 years, we should see this technology become standard on most new laptops. When that happens, I think you’re going to see video conferencing take off. I for one can say that since I got my new MacBook with the iSight camera pre-installed, the cool factor combined with the “well, I really should use this camera I paid for” factor have led me to conference several times with friends — the few of them who also have the MacBook.

The key point here is that including the camera as a default option in mid-range laptops could literally change the face of how we communicate. By just making something “standard,” companies with huge market share like Dell can lower the AE of doing something quite dramatically. Imagine if they included a small drawing tablet, pen, and software with every computer (I obviously in no way advocate this); I think you’d see a lot more doodles floating around. That’s power.
.

Voting

.
Even though this most recent midterm election was considered one of the most important of the last 50 years (with both houses of Congress up for grabs), voter turnout was alarmingly low. The problem is that the AE of voting is alarmingly high.

I find it peculiar that registering for the draft basically happens automatically, and they somehow find you, but registering to vote is a pain in the ass. Sorting that out could be one of the best ways to improve voter turnout. But assuming you do register, the process of voting itself is incredibly complicated and difficult, considering what you’re actually doing. It seems unnatural that we’ve figured out how to conveniently, safely, and with perfect accuracy let people transfer large sums of cash on the side of the street in millions of locations, yet we can’t get what amounts to answering a few multiple choice questions on a screen to work. We’re still required to go to some empty elementary school in the middle of a workday to wait in a really big line. Nobody likes lines, even when there’s a topsy-turvy fantastical ride at the end of them.

So the first step to lowering the AE of voting would be to fix the process. But those aren’t really the biggest impediments. After all, if someone was really passionate about it, that would be a catalyst enough to truck through the procedural annoyances. How do you get past the malaise?

One way is by asking people to vote. There are many “get the vote out” campaigns, but if they’re not dwarfed by attack ads and heated rhetoric, they’re sponsored by some group that at least half the people don’t agree with. A non-partisan, widespread effort to get people off their asses on election day might work wonders. Applying concepts from Web 2.0, a viral campaign to “get one friend to vote with you” might actually work.
.

Google

.
Google uses defaults masterfully. Since their IPO, they’ve extended their reach far beyond search, and it shows. They release new stuff all the time, like Gmail, a news aggregator, a web-based word processor and spreadsheet application, a map/earth visualization system, and more. They develop some of these products themselves, but in large part they buy small start-ups that have already done the development work. The peculiar thing about many of these services is that most people never used them before they were bought by Google. Why?

It’s not like Google makes a massive marketing push each time they release something new. All they do is add a button right above their search bar that links to their latest offering. The services themselves, as part of the original company that developed them, were always easy to use, and for the most part Google doesn’t change the actual product. The thing is that before Google, they required a huge AE to find — you had to read stuff like TechCrunch or the Valley Wire to see that they were released, and so only nerds found them (have you heard of Writely, or Keyhole, or Pyra Labs? Those are Google Docs, Google Earth, and Blogger). Otherwise these small companies would have had to run large ad campaigns, or viral marketing schemes, or other stuff that they clearly didn’t have the money, brains, or time to do, to lower that AE for you. But once they’re on Google, the AE takes a massive dive. Millions of people visit Google every day. When the page changes because of a new product offering, the thing is almost guaranteed to take off.

That’s why I predict that Google Docs and Spreadsheets, which most people don’t even know exists, will become a huge player in the word processing world the minute Google decides to put it next to Maps on their front page. That’s the power of AE.
.

Endnotes

.
[1] Why does Firefox only hold 10% of the browser market when it’s, by any and all accounts, better than IE? Simply put, most people use IE by default because IE comes packaged with Windows, and it’s good enough that switching to Firefox, as easy as it may be, just isn’t worth it. Downloading it, installing it, and learning how to use it, despite being really simple, require too much AE for someone who *just* browses the web, which is most people.

This essay itself was designed AE in mind. I tried to make the first segments the easiest to read, using short sentences and familiar examples (like the first day of class instead of a web service), hopefully giving average reader a chance to get here. If you did make it, maybe I’m on to something.

Digg!

.

Comments

.

Eric B. Nov 22nd, 2007 at 7:47 pm

This principle applies to so much different stuff, as I start to think about it. You actually philosophized and produced something other than just a fluffy word-cloud. More!

.