If you’ve been compulsively refreshing our website lately (and let’s be honest: who doesn’t spend all day compulsively refreshing our website?), you may have noticed that we’ve been experimenting a bit with our home page design. We A/B tested a number of different designs before settling on the one we have now, and we wanted to share a behind-the-scenes peek at our process.
Why A/B Test?
It’s no secret that small changes to a site’s design can have a big impact on conversion rates. One site increased their signups by 30% just by changing the color of their button from red to green! A/B testing—releasing multiple versions of a page into the wild and measuring the results of each—is a great way to find out which potential changes are the most effective.
Our motto is, “You don’t know until you try.” When came up with several different designs for our homepage, we initially got sucked into arguments about which was one was best. (Guess we needed to get our heads out of our bikesheds.) But we quickly realized there was no point in arguing amongst each other. Putting our designs out into the real world and seeing how users responded would be far more valuable.
Our original home page (above) had served us well over the many months it’d been up, so our first inclination was to make only minor changes. We tried two new designs: Version A (below left) added some teaser information about the benefits of PlayThru and callouts to some of the well-known sites that use it, while Version B (below right) was a more simplified take on the same information.
But neither did the trick. Version B had no noticeable impact on our conversion rate, while Version A actually decreased the number of people who signed up! Maybe they were overwhelmed—as 37signals discovered, more information isn’t always better.
We found greater success with our next design, which boosted conversions by around 5%:
This lined up with a lot of what we’d read about landing page optimization—namely, that people love pictures. There was only one problem: that page was ugly! No one wants a cheesy stock photo—especially one of a “businessman”—on their home page.
Then we had an idea: since we’re all about games, why not try some photos of kids playing games? So we tried these two designs next:
I was personally a fan of the one on the left (although Reid thought that the little girl was, quote, “creepy”), but ultimately the design on the right, with the boys playing leapfrog, performed better. It was crowned the victor, and it’s now what all visitors see on our home page.
The lesson? We can spend hours debating which designs will work best, but putting them in front of real people is the only way to know for sure. The results might surprise you!
Last week my partner Tyler and I were in San Francisco. We hadn’t booked a hotel room in advance, so when we got in our car I pulled up Expedia to look for a room. This took a good fifteen minutes. We debated location, did we want to be by a bus stop, was internet included, etc etc. Finally we agreed on a place, and ended up saving about $40 per night.
Today one of our developers shared Parkinson’s Law of Triviality with me. After that hotel debate, it hit home. For him it came up in the context of a forum discussion that had degraded to a formatting argument, but it applies to us constantly at Are You a Human.
It boils down to this: the more complex an issue is, the less time is spent discussing it. Humans tend to focus on the things we understand, rather than the things that are most important. It makes us feel good to give an opinion; being involved in a discussion makes us feel valuable, even if the topic isn’t worth discussing at all. Parkinson’s classic example is a meeting with a nuclear reactor and a bicycle shed on the agenda.
The bike shed is getting most of the discussion. This can be a huge time killer, and we don’t have extra time lying around.
Those fifteen minutes Tyler and I spent debating a hotel could have been spent in a myriad of more productive ways: reviewing the details for the meeting we were driving to, debating how we want to position our company to investors, discussing which of a dozen possible dev projects would have the most impact. Forming an opinion on a hotel was easy, so we did that rather than spending time on the decisions that have huge consequences for the company.
The risk of this is especially huge when you have technical/non-technical discussions like we do every day. I’m obsessed with Are You a Human, and I have an opinion on everything, so when we have product discussions, I’m sure I steer the conversation towards the less technical, likely less important aspects that I understand.
So what can we do about it? Unfortunately there’s probably no trick here, we just need to be conscious of the tendency to invest in the wrong places and watch where our time goes. To make sure we’re constantly thinking about this we’ve added a new Are You a Human code word.
We did a previous post about hot potato and tangerine. They’re simple phrases that we use to keep each other on task and work efficiently. Well, now we’re adding “bikeshed” to our list. The next time we spend more than ten seconds discussing how large the Twitter icon on the blog should be someone can yell “bikeshed” and we can move on to more important things.
In entrepreneurial circles, there’s been a lot of debate about business plan competitions, with wide disagrement about whether they create any value or are worth the time that’s put into them. Entrepreneurs like Steve Blank have even started their own competitions to break out of the traditional “business plan” mold.
Many complain that the current breed of competitions is nothing more than a beauty contest. And to an extent I agree; no fledgling business wants to be known for a good looks and inadequate substance. But as a participant in numerous competitions (including Rice University’s, Wake Forests’ Elevator Competition, Accelerate Michigan, and the Michigan Business Challenge), I have observed three distinct benefits that these competitions provide for entrepreneurs.
The 2011 Are You a Human team holds a giant check from the Rice University Business Plan Competition
- Deadlines. Deadlines drive action. I am constantly pulled in multiple directions (businesses, school, family, church), and dedicating time to work on just one thing is scarce. But if you’re working on a startup in conjunction with several other activities, the structure a deadline provides is essential. Competitions force our team to concentrate our efforts and get stuff done—even pulling all-nighters if needed. Even if you’re working on a startup full-time, a deadline can help add some pressure to doing customer discovery, launching, or getting a pitch ready for investors.
- The lesson that it’s more important to learn than to win. Be forewarned: if winning is your only focus, you might be inclined to leave important information out of your pitch and only present the details that makes you look best. True, those numbers can impress judges and tell a hypnotizing story, and they might even lead you to win some money. But you don’t want to get addicted to big numbers in fancy presentations at the expense of real data. The real value of a business plan competition lies in discovering a viable business model for your idea, or discovering that there isn’t one. Competitions are also an opportunity to build your network, as judges are usually either potential customers or potential investors. The late nights and fierce rivalries of competitions are only a glimpse of the rocky road many startups face as they get going, so take criticism, look at the holes in your vision for the company, and engage with experts in your field.
- Validation. Business plan competitions provide the most value for fledgling businesses that need an opportunity to focus very early on and determine validity. For them, there can be huge value in getting a chunk of change that will allow for some more runway. On the other hand, for businesses that already have funding or a proven market, business plan competitions can be distractions that kills progress. They may be awarded $2,000 or even $1 million, but the time and energy put into the event could have equaled a potential customer or finished prototype worth 10 times that amount. It’s a trade-off, and a balance must be struck.
Yes, business plan competitions can stall progress or be distractions, but they can also provide real rewards. If you’re thinking about entering one, don’t throw away the good with the bad.
Give a businessman a database report and you’ll satiate him for 5 minutes. Teach him how to use a tool to find the information himself, and he’ll be satiated for a lifetime. Well, maybe not that long, but at least you’ll get a chance to write code for a change instead of “SELECT * FROM.”
Much of my time here is spent developing internal tools for other employees to use, or trying to “program myself out of a job.” There is never a shortage of projects to work on. There are more users to manage, more games to create, more customizable functionality to add, and more ways of anaylizing our data every day. And if we’re going to continue to grow our user base, I can’t be writing queries by hand everytime we want to know what account is associated with what email. That’d be absurd.
The tag manager is just one of many beautifully-designed internal tools we use each day, in this case to look up information about customers’ email addresses.
Developing tools for others to use allows them to interact with parts of the system they would otherwise be unfamiliar with. I’ve written tools to manage which customers get served which games, tools to manage what emails we send out, and tools to display hundreds of bananas in a browser. I’d rather not be a code monkey completeing monotonous tasks, so these tools let me move on to more interesting things.
Like watching a video of a man getting hit in the face with a fish.
“Tech support.” Are there any two more dreaded words in the English language? Since the dawn of the computer age, support has been a hated job and the butt of lame jokes. When I was first told that part of my job would involve doing support, I almost vomited. Here are just a few of the pleasures I thought I’d be subjected to:
- Answering the same questions over and over again
- Being “yelled” at (a.k.a. emailed at in all caps)
- Being asked to do impossible things
- Praying for the sweet release of death
Yes, I was decidedly unexcited to start doing support. But it turns out I was completely wrong. Here’s why.
This is the image that came up when I searched for “tech support.” Sadly, no one at Are You a Human is as good-looking as this guy.
1. Support makes people happy.
It’s a surprising paradox that customers often like you more if you successfully help them with a problem than they would if they had never encountered that problem in the first place. It’s all about expectations: we don’t get excited when a product we buy works—after all, that’s what we expect. But we do get excited when we get great support, since we so often expect the opposite. Our customers are often thrilled to get a quick, friendly response from a real person, and it makes me happy to see them happy.
2. Support provides a direct connection to our customers
Doing support is about more than responding to problems on a case-by-case basis. Over time, talking to our customers every day gives me a sense of what parts of PlayThru are working well and what parts can be improved. Vanilla games and BuddyPress support are just two of the features that came about because repeated support requests told us you guys were interested in them.
Plus, it gives our customers a direct connection to us, and that’s an area where we can really shine as a small company. Email most companies and you’ll get a response from a random employee who probably won’t know—or care—much about your problem. It’s different with us. Have a problem with the WordPress plugin? There’s a good chance the guy who wrote that plugin will be the one to email you back.
3. People are nice!
When I first started doing support, I expected that people would be angry. We all get pissed off when technology gives us problems. And sure, some people do write in using language that’s unprintable on this blog. (A personal pet peeve: when people send angry support requests with fake emails, like email@example.com. Why waste your time and mine writing in to support if you’re going to make it impossible for me to respond?)
But for the most part, the people I talk to are all extremely patient and gracious. Would it be going too far to say that doing support has restore my faith in humanity? Yeah, that probably would be going too far. But still.
So there you have it: the amazing true story of someone who actually likes doing support. The next time you need tech support—from anyone, Human or not—don’t forget that there’s a real person on the other end of the computer. Oh, and the first person to write into our support telling me that they’ve read this blog post will get a special prize. Let the games begin!