7 Tips To Hosting an Awesome Hackathon

No Comments

So you are thinking of hosting a hackathon but you’ve never done it before or could use some tips. I’ve been a part of many hackathons as both participant, spectator and judge. Some hackathons were great, some were so so and some were just awful. Here are some things to consider when holding a hackathon.

  1. Connectivity One thing we NEVER want to worry about at a hackathon is connectivity and bandwidth. That’s a given. We should only have to worry about our hackathon project, not trying to find a wifi signal or futzing around with the Wifi password. Along with that, there should be plenty of outlets. We really hate going to hackathons and having to fight for power outlets.
  2. Clear schedule / agenda We want to know when things begin, when code freeze occurs and judging takes place. When lunch and dinner is served is also very important.
  3. Prizes What’s going to motivate a developer to give up his weekend to code on your API? I would venture developers who come to hackathons already have side projects. They probably command a decent salary or consulting rate. Prizes have to be compelling or worthwhile. Having multiple prizes also helps so it’s not a winner take all situation.
  4. Multiple categories/judging criteria Having multiple categories for prizes is another great motivator. There could be different prizes for Best Overall, Best Design, Best Mobile, etc etc. This way someone might feel like they might have a shot at winning if one of the categories matches their strong suit.
  5. Make sure it works If you are holding a hackathon where developers use your API or code library works. The hackathon is NOT the place to be debugging it. The developers are not there to find your bugs. It should just work. Along with that, make sure that the API key and whatnot are readily available.
  6. Good food Pizza and red bull are great but many people I run into have special dietary needs. Some are vegan, some are vegetarian. Often times these developers are totally neglected when it comes to their dietary restrictions. Make sure you take care of these folks. Make sure you also have enough food. On especially long hackathons, snacks are always welcome.
  7. Clear/accessible location People hate getting lost. Make sure to provide adequate instructions to finding the location. It also helps to have signs saying “Hackathon” pointing in the right direction. If the location is locked down because it’s the weekend (as hackathons are often held on weekends), then make sure that there’s someone to let people in.

Popularity: 9% [?]

ShelfLuv PostMortem


ShelfLuv launched over two years ago and I’m sad to say the time has come to shut the service down. It was a very simple idea, to be a better way to search for books. I was a very voracious reader then and wanted to create a very nice interface to find and show off the books I’ve read. A little later (like 6 months), I revamped the site and made it so you could not only create a page you could share of your books but could view a feed of people’s reading activities.

It had some great coverage early on. The site looked nice. I was pretty proud of it from a design perspective, but there were many problems with it. There was no real traction or usage. It was probably up for longer than it needed to be, mostly because I was too lazy to shut it down. It was easier to just do nothing.
However not all is lost. I learned some great things from it.

Final Stats

  • Total Registered Users: 3,290
  • Bookshelves Created: 712
  • Books Added to Shelves: 30,046
  • Books Searched: 1,209,627
  • Comments Created: 1,345
  • Users Followed: 1,573

We used Facebook registration to simplify the onboarding. That probably was both good and bad. Some people do not like to use their Facebook account to sign up for services.

Users added about 10 books to their shelves on average. In reality it was a bit more skewed. I myself had over 300 books added. Some users had even more than that. Basically it was one of those 80/20 rules. Most users signed up and did very little while others got it and got into it.

One of the right things I did was letting people search without having to have an account. As you can see, people did search for books. People weren’t required to login or register to immediately use the beautiful infinite scroll instant search. Also, because we made it “instant”, it would do the searches as you typed, thus lowering the barrier to searching even more.

So what went wrong? If I had to pick one thing it would be we moved too slowly. It took a long time between just having instant search and launching the profile features and feed. We didn’t iterate fast enough. There were many reasons for that. For one I didn’t have much in terms of tech. Although I did all the front end coding and design, I didn’t have a full-stack developer on my team. Out-sourcing that took time and money.

The other thing that was obvious in hindsight was that people just don’t consume books as much or as fast. Not as many people consume books compared to people who post photos for example. The turn around time to consume “content” and produce a piece of content is just long. If I had come up with ways for people to engage with the site other than adding books and commenting, people might have interacted more.

Popularity: 7% [?]

Your App That Nobody Uses Doesn’t Need a New Design

No Comments

Every so often I get approached by someone who wants me to redesign their app / website thinking that will be the solution to their problems. They’ve gotten as far as getting their idea realized, launched and it’s tumbleweeds. Nobody uses the app; nobody cares. They get frustrated and think a fresh new look will draw in the crowd. More often than not, it’s mostly a cosmetic redesign – or pretty-ing it up. I hate to break it to these people, but for most cases, a new design will not do much.

The reason it will not do much is because usually the problem doesn’t have anything to do with whether or not people think the app looks pretty or not. It’s usually a more fundamental problem than that. If you were to judge a site like craigslist, reddit or lolcats by their design, they would fare pretty poorly but they are all top sites.

When nobody uses an app, it comes down to several issues, none of which have to do with design. The number one reason is it doesn’t sufficiently solve a problem. Maybe the problem isn’t big enough or painful enough. Maybe the app doesn’t provide enough delight. Also it could be that the app isn’t that big an improvement over existing solutions. Once people get into a habit of using something, it’s really hard to dislodge that incumbent unless it’s significantly better.

Another reason is maybe not enough people know about your product. A fresh new design will not help with that either. You actually have a marketing problem, not a design problem. There are so many web and iphone apps out there now that it’s just noise to the common person. There are actually over 500,000 apps in the iPhone App store right now. That makes it really hard for your app to rise above the rest.

If you get past all this and still nobody uses your app, maybe you do have a design problem and a fresh coat of paint isn’t going to help either. Maybe using your app is too painful, meaning there are too many steps, too much friction. By that I mean maybe your app makes the user jump through hoops to derive value whether it’s to produce content or consume content. In this case, in order to solve the problem, we really need to understand user behavior first and foremost. What are users actually doing on the site or app? Where are they giving up on your app and dropping off? If we don’t understand and know this for a fact, we really don’t know what we are solving for or how to fix the problem. What you need here, is not a new “design” but data and lots of it.

One of the first things I recommend doing is gathering intel. Figure out what users are doing. You can do this several ways. Install a tool like Olark and talk to your users. You will learn so much. Pain points will crop up over and over. Observe your users. Do this by simply asking people to test your software and stand behind them watching what they do. Resist the urge to guide the users or tell them what to do. Let them figure it out or struggle. Look at the analytics. See where people are leaving your page / site. Use tools to figure out if people are performing the desired behavior on the page they are on. If people leave, figure out WHY they left. Did they find what they needed and leave? Did they give up in frustration? Did they find out that this wasn’t what they were looking for?

Let’s illustrate some concepts with an example. I’ve changed the topic of the site to protect the innocent. There’s a certain recipe site a client wants to improve. People don’t seem to be spending too much time on the site. People browse recipes, find something and leave. Looking at the recipe page itself, it’s easy to see why. They have one call to action – add recipe to My Recipes. That’s it. People look for something they want, find it, get what they need and leave. One of the first thing I suggested is to take a page off Youtube. On the right hand side, they have recommend videos or related videos. Perhaps the recipe site should include recipes that go with the recipe they are on. Or maybe show other recipes based on similar ingredients on the side. That way, when users come in, see what they want, they might see something else they might like.

Another reason why the recipe site is languishing could be because they also only have a few recipes. Maybe people come and cannot find what they want and leave. There’s just not a lot of content there. It’s like going to a party and seeing nobody there. A redesign would make it much worse – now you have a fancy party that looks dead. Look at the analytics; what happens when they search? Do they search and leave? Do they search and try to look through all the search results? In this case, their efforts would be better spent seeding the site with more recipes.

If you require your users to sign up, how many are abandoning the sign up process? Perhaps you need to rethink the sign up form. Figure out what’s the bare minimum information you need to create an account and then let the user play with your app ASAP. Maybe you can even switch it around and let the user play with your app before ever needing to create an account.

In short, when wanting a new design, it’s important to understand why you need that new design and whether or not that will actually solve your problem. The solution could be as simple as moving things around or adding some relevant content to keep the user engaged.

Popularity: 7% [?]

Grads: Want a job? Do a project.

No Comments

So you just graduated and are looking for a job. The problem is your classmates just graduated as well and they are just as qualified as you if not more so and they are also looking. What do you? If you really want to show your employers what you are capable of, you create a project for yourself. If you are a writer, start writing; if you are an illustrator, start illustrating, if you are a designer, start designing; if you are an MBA – well I don’t have all the answers.

The point is instead of talking about what you can do on your resume, just show people what you do. Of course this is going to be a problem for people who can talk a big game but can’t deliver the goods, but that’s the point. You are trying to separate yourself from that crowd.

Every time I launch a new project, I always get a flurry of job offers. When you take on a project, that is infinitely more tangible than a line in the resume. If a picture is worth a thousand words, a project’s gotta be worth more than that. Employers can see what you have to offer instead of having to infer it from a piece of paper, among a huge pile of papers.

So I tell you, recent grads, if you want a job, starting putting out work.

Popularity: 7% [?]

What Do People Do On Pinstagram?

No Comments

One of the coolest things about having a popular site like Pinstagram is the ability to get meaningful data about user behavior. We integrated MixPanel to track every type of user activity including any clicks, and scrolls. Without getting into actual numbers, but rather percentages, we can see some really cool user behavior.

  • 43% of all activities are scrolls
  • 16% of activities are zooms
  • 15% of activities are tab clicks
  • 12% of activities are likes
  • 1.3% of activities are comments
  • 1.7% of activities are searches

Another great data point is that our users average about 10 scrolls each. By scroll I mean, they reached the end of the page, and we present them with another set of photos. That’s equivalent to 10 pages views. I doubt that if instead of an infinite scroll UI, we had next page buttons, that we would have had 10 page views per user. What this means is that the infinite scroll paradigm is really low friction. People don’t mind scrolling down, but pressing buttons is somehow more work.

Likes make up 12% of all activities. But comments make up a measly 1 percent. The ratio makes sense to me since Liking requires a click and requires some effort compared to just scrolling. Commenting requires even more effort and thus is an order of magnitude less than Liking.

Popularity: 8% [?]

Older Entries