Nine APIs, 100 Sandwiches, and Countless Dog Selfies

A Hackathon Tale

Pictured here: extremely photogenic pup (Hi Thomas!). Not pictured here: massive bribing with liver treats.

Originally posted on my now-defunct personal Tumblr in December 2014.

was 9am on Saturday morning. I had a Nutella portable pack in one hand and a cream cheese-smeared bagel in the other. My laptop was humming. This was going to be a long, spectacular day.

Where were we? Aviary HQ at Cooper Square. But let me backtrack a bit.

Last week, I signed up for Photo Hack Day NYC as a Designer. I would be on a team with former and current General Assembly students: Brian, a UX Designer, Jackson, a 7th week Web Dev immersive student, Denise, an 11th week Web Dev immersive student, and Ben, a Product Manager. For most of us, this was our first hackathon, and the first time we would all be seriously collaborating with cross-functional teammates.

I must admit that although I knew the basics (36 hours, a photo theme, nine APIs, competitors), I didn’t exactly know what I was getting myself into. I always imagined hackathons as an unreachable fairy tale competition of superbrains, involving copious amounts of pizza, keyboard typing punctuated with the occasional grunt, bros fist-bumping each other while double-fisting Red Bulls — all the while death-eye staring everyone else in the room.

Thank god I was totally wrong. So how did it go? Let’s see.

The Idea

I had this idea pre-hackathon while furiously brainstorming something to design. I had been browsing Instagram absentmindedly, and I happened to see this random Discovery page. How many dogs do you see out of how many photos?

The answer is 8 out of 12. Don’t forget the photo of the dog with the picture of the dog in the upper right corner. Also, don’t quote me. Math is not my strong point.

Long story short, I had a huge realization. Dog photography is huge. For the super dog accounts on Instagram with thousands or even millions of followers, owners usually have DSLRs and sophisticated treat gadgetry to make sure their pictures are perfect. However, for the average dog owner who just wants a great photo to post on social media, it is really tough to get your dog to stay still! In my case, I usually use treats to get my hyper dog’s attention, but it is always a really awkward balancing act. The idea I pitched to my team was to use the one thing (other than treats) that gets their attention — sounds! A half-second before the camera clicks, there would be a pre-programmed list of sounds you can choose from that would make your dog look intently at the camera. This seemed like a good insight to hack around.

But first, I needed to quickly and cheaply test an important part of my hypothesis. Before the night of the hackathon, phone camera in hand, I took some shots of Thomas while vocalizing a variety of weird sounds. I think in these shots I was barking at him. He usually zips around the room, tri-colored blur, so getting him to stand still for these shots is amazing. Seriously, the head-cocking has never even been photographed before.

First Steps

At Aviary, after gorging ourselves on a delicious breakfast (RIP Nutella jar, you were loved) we watched CEO of Aviary Tobias Peggs kick off the hackathon! I really wish I had a photo of the host, Emily, who had us cracking up the entire time.

We were then treated to awesome demos of all available APIs by sponsors. After the presentations, we had a group pow-wow to discuss our idea, brainstorm possible features, and narrow down the path to get to a workable MVP. Ideas were thrown around that we had to validate. This is where we split up.

Ben (PM) and the two devs worked on elevator pitches and user stories.

Brian and I would be assumption testers — given the small amount of time, we decided on just sticking to one tool in our UX kit — interviews. With this in mind, we jotted down our assumptions:

  • People, especially dog owners, take photos of dogs.
  • People share these photos.
  • People take close-up photos of dogs.
  • People have a hard time getting their dogs to stay still for a photo.
  • Barking sounds, meowing sounds and treat bag sounds get the attention of dogs.
  • People like to edit their photos.
  • People geotag their photo locations.
  • People use hashtags.
  • People print photos of dogs.

With these, we quickly thought up a general guideline for our interviews, and within 10 minutes we were up and stalking unwitting bystanders.

Given the short amount of time our target was to interview 5 people, and we were able to do a few more than that. Take these findings with a grain of salt because we are interviewing a bunch of people in the tech photo-industry, but here are our findings:

  • People, even non-dog owners, enjoy taking photos of dogs, and frequently do.
  • People share these photos via Facebook and Instagram, and through text and email.
  • Majority of those interviewed take close-up shots, with 1 or 2 preferring action shots. Interviewees really like seeing expressions.
  • Most people have a tough time getting the attention of dogs for photos. A lot of them use verbal cues (calling their name in a baby voice, treats, or a click or whistle).
  • People like editing photos, and really prefer one-click sharing.
  • People geotag their photos.
  • People use hashtags, though some of them ironically.
  • People were split on photo printing.

Because we took a sampling of just a few dog owners and mostly dog-lovers, and because we obviously didn’t have dogs on hand, we were unable to test the interest of (other) dogs in barking or meowing or treat bag sounds. The weird and awkward barking exercise I did last night with Thomas would have to stand for hypothesis-confirmation for all dogs. It’s okay though, I think Thomas is always up for hypothesis-testing.

Narrowing Down to an MVP

With all these findings in place, we met back with the team, and by this time they had drilled down to key elevator pitches and user stories. Combining our research and their work, we were able to narrow down to what we assumed was a workable MVP. The surprising finding during our research was that people relied on their own brand of verbal cues when getting dogs’ attentions, which led to us including the ability to record your own sound. We would do a web app (our developers only learned web app coding during their immersive) with the ability to choose between pre-recorded sounds and the ability to record and play one of your own sounds, it would allow for photo taking, it would allow for camera flipping, it would not include flash capabilities (simply because of dog eye biology which bounces light off and makes them look like robots when you turn flash on), it would allow for photo gallery access, and it would allow for editing with the Aviary API. Ideally, it would also show an on-boarding process.

Even though geotagging, automatic hashtagging and photo printing APIs would have been awesome, it was deemed too complicated and iceboxed. Another key insight that we got from our research with users is that blurriness is an issue with dogs, so burst mode would be really really important in a full product. Again, in the interest of time, we relegated that to icebox as well.


Brian and I sketched a few wireframes very quickly, and then I started building the mock-ups in Sketch. We did usability testing on it as we went along, and made a few tweaks here and there. On its final iteration, this is what we came up with:

(I knew my numerous dog selfies would come in handy some day!)

Quickly churning this out allowed me to show it to the devs, who were at the time researching the key parts of the planned functionality.

Zoom forward a few hours, and it was 10pm at night. We had very very basic functionality. Everyone was exhausted. Devs were buried in their computers trying to figure out how to save a taken photo for further editing. The UX Designers and Product Manager were about to call it a night. THEN Ben asks an innocent question to Jackson and Denise about their progress on adding the camera sound, and the devs bolt upright. They had completely forgotten that they needed to building a camera app, and not just a laptop webcam trigger! (this is what happens after a very long day) They definitely didn’t know how to make a camera app, as that is quite complicated. We were stumped. Things looked really scary. Slightly sad and dejected, we decided to call it a night and start fresh the next day, and I quickly made my mock-ups into a prototype using Pop just so we had something to present for tomorrow. Jackson was a champ and stayed overnight to try and figure stuff out.

Sunday morning breakthrough

Sunday, the day of the presentations. Running on no sleep at ALL, Jackson had an overnight breakthrough! He can now take a photo online to the looping pre-recorded sound of me whistling and calling Thomas’ name, and then, confusingly, he can look at the source, pull the photo URL, and then plug that into the Aviary API and edit the photo from there. Hmm… not a super smooth process, but we decide that it would have to do. Jackson and Denise keep working through the morning, and with Conor’s guidance, manage to come up with a few more breakthroughs! We all get extremely excited.

At noon, our live demo guy arrives! Here he is, looking like he is only interested in live demo-ing a Sleepy’s mattress.

In the meantime, Ben takes the lead on planning the presentation. It will require an intricate dance of pitch, prototype, Jackson on the keyboard, and dog selfie. We practice it a few times, sans dog. I get nervous about Thomas getting desensitized to the sound. I polish up the prototype. Everyone munches through lunch with bated breath.

Presentation time + Murphy’s Law (of course)

Woofer is going last! We watch the presentations, and each app is so amazing. It is definitely a lot of pressure. When we go, things get a little hairy. Jackson’s laptop won’t connect to the display system, so we improvise and face the cellphone display camera at it, but the camera won’t focus and the screen is too bright to see any text. We nervously giggle our way through the presentation. When it is dog selfie time, Thomas is totally desensitized to my whistle and ignores it, making our selfie a blur, which in turn makes the audience roar with laughter.

And I realize then and there, even if things are not perfect, that it’s okay. This is wildly fun and I love it. I especially love how supportive everyone in the crowd was of each other. It wasn’t the crazy bro-fest competition I had originally envisioned.

After deliberations, they started announcing winners. Aaaand… we won!!! First hackathon ever and we actually won! Our scraggly team took home Crowd Favorite, Honorable Mention and Best for Photo Creation! Here is our victory shot.

There may or may not have been some victory dancing involved post-photo. Honestly, these wins were probably 80% due to Thomas, so we made sure to give him plenty of hugs and belly rubs in addition to all the hugs and belly rubs he’d been getting throughout the day anyway. Basically, it was a good day for the dog.

As a recap: in 36 hours during a rainy weekend, we sat at a table, ate too much food, and… designed and built something out of nothing. We collaborated as a cross-functional team, validated assumptions, crafted an MVP, and did a good job communicating ideas to developers and altering them based on realistic timelines. I was able to polish up my wireframing and prototyping skills. We didn’t do everything perfectly, but we learned from our failures enough to keep from making the same mistake twice. If this is what hackathons are like, then I can’t wait for my next one.

But first…

Hackathon pro tips (for next time)

Do not keep water near the computers. This we learned painfully, when one of the water glasses spilled onto Ben’s laptop. It fried his laptop. We donated some of our prize money to help him get a new motherboard (or something). The lesson is: Don’t drink and hack, friends.

Don’t be afraid to ask for help. When Jackson and Denise hit patchy spots, they asked for help. Conor from Aviary was awesome, and really helped us connect the dots.

Find a team on-site. I believe that hackathons are all about collaboration first, and second about making cool stuff. Working with a group of friends is awesome, but nothing compares to the learning experience working with strangers would provide.

Who needs pizza? …when you have breakfast buffets, Mediterranean sandwiches, Korean fried chicken, bacon galore, etc. Aviary really hooked us up, in more ways then one. If there are any Aviary folks reading this, THANK YOU.

Bettina E. Bergendahl

Written by

UX designer at Belkin currently making WiFi less scary. Has a dog named Thomas who nips knees, and a new baby who nips Thomas.