Seeing this laid out again after the prototype gave me a completely new perspective on the user flow. Though people were using the site more or less exactly how I designed it, two things immediately stood out to me:

Assymmetry

Overall, it’s a pretty lopsided and, as a result, inefficient process. All the decision-making is accorded to the first user, while the second user can undo all those efforts with a single click. Ironically, I had designed it this way to make it more efficient… I had wanted to avoid a protracted discussion on ideal cooking times that leads nowhere, resulting in cooking that never happens. However in reality, people had this discussion anyway, outside of Hotpot’s purview. And forcing people to send all those ‘concrete’ invites resulted in a lot of wasted back-and-forths. Some even abandoned the process after a it resulted in a “maybe” the first time.

Alternative Channels of Communication

One new development was the blue circle: people needing to nudge each other into action because quite often the invitation email went ignored. I hadn’t quite expected that this would play such a big row, but in hindsight, the necessity for that nudge makes a lot of sense. We get so much email these days, it’s common practice to bypass an email from an unfamiliar service or website.

Looking at this, I realized now that I made the classic mistake of being an overzealous designer, trying to design everything. Instead of recognizing that Hotpot exists within a crowded network of services used by a single person daily, I had tried to make Hotpot take ownership of almost the entire communication process. Rather than allowing friends to communicate in whatever way was most natural and effective for them, I was asking them to use this specialized invitation which lacked the social capital to attract attention.

The “formal invitation” was one of my early darlings going all the way back to October. It was spurred on by a lot of positive feedback, supporting the idea that it helps make the cooking event feel even more special. Now I know this is not exactly true: an invitation email, no matter how well-designed, is not going to feel special in a crowded inbox.

Again, this is exactly why you make a live prototype.

So what would an improved process look like? Here’s one take:

In this version, I aimed for a few things:

• Earlier discovery for User 2: Earlier discovery means a longer time to get acquainted with Hotpot, rather than being taken by surprise with the invitation.
• Increased symmetry: Decision-making for the cooking event now happens together. This allows the 2nd user to become invested in the invitation process, rather being confronted with basically an ultimatum: “I already set the time and date, so take it or leave it.”  Friends that plan something together are more likely to actually do it together.
• Take advantage of existing channels: This version offloads the hard work of reaching consensus to whatever communication channels work best for the users. Doing so gives users more flexibility and control over that conversation, and allows them to take advantage of existing social capital to ensure  responsiveness. The downside is that Hotpot plays less of an intervening role, so this wouldn’t work between people who aren’t well-acquainted to begin with. If Hotpot evolves into a “meeting people” platform, the service will need to play a bigger mediating role. But for the purposes of getting a minimum unit of value working by the end of thesis time, I think it’s a good move.

Doing this was extremely useful in helping me to imagine the next steps. I admit to feeling rather lost and helpless last week, as I couldn’t quite figure out why people weren’t using this thing I worked so hard to make. Now, I have some concrete clues as to what to do in the remaining weeks.

I think that, rather than adding more features or imagining a fancy future for the product, I would rather make improvements to the existing prototype. My biggest hope is, by the time thesis is over, people will actually use this to cook with their friends. Rather than attempt to inspire and delight my thesis audience with visions of a beautiful imagined future, I want to inspire and delight with the actual experience of cooking with your friends. That alone, I think, will have enough beauty to fill an auditorium.

There were both successes and failures. First, the successes:

• People responded very positively to the playful visual design, including several fully grown men with luxuriant beards. I win!
• People largely understood the core concept and expressed enthusiasm for it.
• For the most part, people had little trouble completing key tasks when asked to: sending an invite, viewing one, responding to one, navigating to a cooking room, etc.
• The code I wrote didn’t completely splatter everywhere as soon as someone tried using it.

But there were some failures, too:

• The biggest failing was that, despite expressing enthusiasm towards the concept (“I want to try this!”), few people actually got as far as sending an invitation, and fewer still made it to the cooking stage. As I soon realized, it’s a lot to expect, but there may also be some flaws in the userflow. I’ll write more about this in the next post, as well as make some guesses as to what needs to be done to fix it.
• One or two people had some confusion over whether the cooking happens concurrently, so I need to be a little more explicit about the “real time” aspect.
• OpenTok sound quality was terrible for the first live cooking session done by independent testers (But to the testers’ credit, they figured out a way around it by just calling each other on the phone while cooking).

There were a host of small bugs too. Those are to be expected, and are relatively easy to fix.

Overall, actually seeing Hotpot live in the real world has been completely eye-opening. As I alluded to in my last post, you just can’t argue with a live prototype. It shows you how things really are, regardless of what you steadfastly believed in design fiction land. The real challenge now is how to interpret and implement changes by the thesis deadline, which is fast approaching. That will be the ultimate game of prioritization.

1. Aim to run 2 prototypes a week (iterate faster!)
2. Experiment with different social groupings (# of people, relationship to host, etc.)

With #1 in mind, I ran another Google+ Hangouts prototype on Wednesday. With #2 in mind, I invited just one person: a friend on the west coast.

Two unexpected things happened after the Hangout started. First, my friend’s significant other arrived on the scene, and secondthe recipe I chose turned out to require a lot of diced onions. The onions are relevant, because my significant other (Yang) is a champion-grade onion chopper (whereas it tends to be an exercise in pain management for me) so he, too, joined us. What started off as a meetup between two old friends turned into a couples cook-off.

First, it entirely changed the social dynamic. Having some strangers in the mix meant there was a lot of mutual curiosity. We asked each other about jobs, interests, what-the-hell-my-thesis-is, and all sorts of other things. Cooking took tons longer than expected as a result. But we had a great time getting to know each other.

Second, and this is the great part, the mutual curiosity turned into an actual dinner together. We ended up finishing the dish at the same time and decided, why not, let’s just stay on Google+. We set our laptop on our table, and they on theirs, and we ate the same dish (albeit cooked 2 different ways) while connected via video chat.

The view from across the table

Unlike while cooking, we were able to spend time looking at one another’s video streams, literally spending face-to-face time together. We had real conversations, without the distraction of simultaneously monitoring a bubbling pot. We clinked beer bottles/wine glasses via our screens, giggled hilariously, and even fork-jousted (no fear of exchanging germs here). It felt a lot like dining with friends in real life, except with a 2500-mile and 3-hour gap in space and time. But by the end of the meal that was hardly noticeable; we were all very happy and full. Prototype #2 turned into an experience that was surprisingly very, very good.

After last week’s prototype, one of my testers sent me the following feedback:

…I would also seriously enjoy to do this with my friends back in Norway on the weekends – so that I could make a hearty lunch while they prepare their dinner (6 hours time difference)

This observation makes so much sense now. I hadn’t thought too much about this before, but food really does have amazing connective properties—not just across cultures, but across space and time as well. After all, this was a friend I hadn’t seriously spent time with in years, and I had just visited his home, gotten to know his significant other, lectured him on knife position (it was seriously worrying), and shared a great meal with him. The time difference can be an issue, as my tester pointed out above, but all that melts away when everyone is affixed at their table with a nice plate of delicious food.

I have a clearer picture in my mind now of what I need to do to recreate this great experience for others:

• make it ridiculously easy to set and confirm a cooking date (across time differences)
• then create an interface that puts food front-and-center when you’re creating it…
• …and melts away into just you and your friend(s) when you’re done cooking and eating together

Image courtesy of @rachelgeylin

The entire test ran from 8pm to roughly 10pm, taking somewhat longer than I had expected. Reasons for this included troubles with the technology, inaccuracies in the recipe itself, and differences in cooking styles (one group doubled the recipe, thus taking longer to chop and prepare the ingredients).

Some notes on the findings:

Invitations & Ramp-Up

On Tuesday, I sent out an invitation with a set date and time to 10 friends. About half responded “yes,” and two more said they would like to come but unfortunately couldn’t make it. The reception was overwhelmingly positive, a good sign!

The invitation included a description of what was to occur, an ingredient list and a link to the recipe itself. I purposefully chose a recipe that was very season-appropriate. There was little to no confusion about how this was going to work— I think everyone was fairly familiar with the idea of Google+ Hangouts / video chat.

The day before, I also sent out a shorter reminder email that included the ingredients list again.

Technology

On the big day, there was some fumbling with the video chat technology early on in the process, but by 8:10 we were all present and introducing ourselves. The Google+ Hangouts interface is fairly easy to use. The only difficulty came when some of us dropped off and had to be reconnected to the same hangout.

A nice feature of the Google+ hangout is that it takes away the awkwardness of everyone having to learn everyone else’s names at the beginning. You can mouse over the video thumbnails at any time to get that user’s name.

The video was choppy and sometimes would freeze, but the sound quality was very good. I found the video quality not to be too much of a problem since most of the time you’re looking down at the prep surface. I only looked up at the video during breaks, or when someone asked me a question directly.

Side note: I also covered my laptop with plastic wrap beforehand. It works really well! You can still type with ease and use the trackpad, but the keyboard becomes essentially waterproof. However, as it turned out, I did not find myself typing much at all. My hands were far too busy chopping, stirring, and sautéeing!

Cooking Process

We kept to a fairly unstructured format, allowing each participant to go at their own pace. The biggest difference in prep styles came in chopping the butternut squash. Some bought pre-cut squash, some had already cut theirs before the prototype started, and others (me included!) had to clean, peel and chop it. There was about a 35-minute difference between those who were cooking fastest and those cooking slowest. There was also much deviation from the recipe itself.

The difference in cooking times didn’t seem to affect the experience negatively, as it then became possible to ask one another “How did you do this previous step x?” Though, towards the end, I got a little impatient waiting for my soup to finish as some of my friends had already started eating. They ate with us, observing us and offering tips along the way.

Communication

We talked mostly about the cooking process—asking each other for tips, whether one deviated from the recipe, etc. There were some side conversations here and there about weekend outings. (We also got to say hi to my friend’s dog on camera!) In the beginning, we also discussed ingredients (especially the squash, and whether buying pre-cut constituted “cheating”).

Communication was exclusively via video, and for the most part audio-only would have sufficed. I had hypothesized that we would be using text chat to complement the video, but we didn’t at all. Voice-chatting is much easier when you have your hands full dicing squash.

One problem with communication was that, with 7ish people total, there was 1) a lot of background noise from cooking and 2) it was sometimes hard to tell who was saying what or talking to whom. However we worked out a system of addressing each other by name and speaking extra loudly when we wanted each other’s attention.

Another issue was that it was difficult to have a good camera angle for some of us. In small apartments with limited counter space, there isn’t always a good place to put the laptop. Some of us went for face-only shots, others (me included) tried to point the camera at the entire prep surface so you could see what was going on. This for me meant putting the laptop on top of the fridge. While this afforded everyone a good view of my kitchen, it was hard for me to see the small video thumbnails along the bottom. The laptop was too far away, and the angle of the screen (half-closed) cut off the stuff at the bottom. These issues should all be addressed in the final interface design.

Interface

The “interface” consisted of a Google+ Hangouts window and a recipe zoomed in to 100% open in another tab. This required tabbing back and forth to switch between viewing the video and viewing the steps of the recipe. Those with a bigger screen or multiple screens were able to have the recipe open on one and Hangouts in another. This is more of an ideal setup, I think.

Conclusion and Next Steps

Cooking together is fun! This prototype has definitely helped give me the green light on moving forward. Now that I know that it’s not only possible, but actually enjoyable to have online “cooking dates” together, I can begin to address all of the problems we discovered above in earnest. I know that some of this will be bounded by the constraints of present-day technology, but many of the problems we identified are solvable through careful interface design.

The next step is to research technologies/tools for building a custom interface. I’ll continue to host Google+ Hangouts cook-alongs at least once every week to test different group configurations (size, relationships to one another, etc.).

To prototypers: Thanks, everyone, for helping me prototype! If you have anything to add, please use the comments section below. I also look forward to your survey responses (hint hint)

I finished Recipe for America by Jill Richardson this week. It was an excellent and comprehensive introduction to all the major issues plaguing the US food system today, covering everything from farming methods to legislation.

Admittedly, after reading this book, I felt overwhelmed (and more than a little depressed) by all that’s broken with our food system. I imagine this feeling is shared by pretty much anyone who has initiated a thorough investigation into a problem, so I’ll take comfort in the knowledge that, if it doesn’t feel overwhelming, perhaps I haven’t looked deeply enough yet. So there, at least I’ve begun to look deeply.

Here are some of my thoughts after reading this book:

Availability precedes consumer choice

One thing I was surprised to learn about was the prevalence of “food deserts” in this country. If you live in one such “desert,” you literally have no access to fresh and nutritious produce, unless you are willing to dedicate a day trip or more to grocery shopping. (The alternative is, you’ll find yourself subsisting on gas station offerings like Coke and Ding-Dongs.)

Amidst all this recent talk about obesity and the wasteful excesses of our food system, it’s hard to believe that such areas can exist in the US, but they do. This just highlights the importance of availability in shaping consumer choices. Fixing the broken food system is not simply a question of whether we can convince everybody to make better food choices; it’s also a question of whether they have the ability to make those choices in the first place. To this end, much work should be done with directing food flows to where it’s most needed, or even encouraging small-scale local agriculture.

Material realities determine relationship to food (know your audience)

The fact that one’s financial situation affects what one eats goes without saying. But, as I found out, not having enough money not only limits what one can afford, it also limits what one can do with what they buy:

“…when you suggest that people buy rice, pasta, and beans you presuppose that they have resources for capital investment for future meals … a kitchen, pots, pans, utensils, gas, electricity, a refrigerator, a home with rent paid, the time to cook. Those healthy rice and beans can take hours … buying a doughnut for dinner does not involve any of those middleclass resources. You pay 55 cents for this meal only and there you are.”

This idea that lower-income people don’t cook seems to run counter to intuition, as we all know from living in cities that eating out is vastly more expensive than cooking one’s own meals. But as it turns out, cooking requires material resources that many of us take for granted. When you are truly living on the edge of poverty, you eat what fills you, whether it looks like a meal or not.

Clearly, one should not always base one’s assumptions on intuition alone. For my thesis, I had hoped that cooking could help people develop a tighter kinship with their food, or at least a stronger awareness. But if many people in the US can’t cook, then the impact of advocacy-through-cooking is limited. That said, there are ways to still introduce stricken communities to the joys of cooking. Community kitchens and potlucks to mind. Either way, this just demonstrates that there is not a single solution to this vast problem. Solutions must be carefully adapted to a particular audience, which you should learn everything you can about.

The government is involved—a lot

Knowing is caring, and reading this book made me actually want to be an food activist. (This lasted about 5 minutes) As I learned, the US government is in many ways the final determinant for what one can or cannot do in the food world: what you’re allowed to sell, where you can sell it, and how it can be produced and processed.

For instance, the federal government legislates everything from the price that is paid for corn (Farm Bill) to what children are fed in public schools (USDA). Laws and court rulings determine what kind of pesticides are okay to use, and even whether one can successfully sue for damaging use of pesticides. The government can even limit the size of an independently owned farm, and whether a small producer can sell to certain constituencies (out-of-state, wholesale, retail outlets, etc.)

While all this sounds potentially helpful if done in the right spirit, US food policy currently acts as a complex support structure for a single idea of how to produce food: a mega-consolidated, industrial model. This model is founded on ideas of efficiency, economies of scale, and profitability; sadly, it does not concern itself with the intricacies of human or animal rights, environmental or community health, ethics, and long-term sustainability. Nevertheless, this is the system the US government approves of, because it provides more food and more jobs at lower prices than ever before in history. Because this is the dominant, government-supported model, it can make it very difficult for a small business to take off. For instance, they are sometimes forced to fulfill requirements that might work well for a large factory farm with economies of scale, but are onerous and unnecessary for small ones.

As much as I believe in bottom-up, consumer-driven change, this book makes the argument that any bottom-up endeavor must be complemented with policy change from the top. No matter how many of us want small local farms and independent producers, if it is difficult to start a farm in the first place, we won’t get those things. Both fights must be fought; one does not supersede or invalidate the other. Indeed, Richardson writes:

I believe that lack of popular outcry is why corporations have come to rule our food system, and with our urging, the government can work with us to take that system back.

The food system is MAD complex

The old adage “the more you learn, the less you know” applies here. I thought I knew enough about how food works in the US after reading The Omnivore’s Dilemma. Not a chance—Michael Pollan barely touches on all the legislative shenanigans surrounding food. And the class differences, especially the plight of the poor (MP is pretty much your average white middle-class dude). And the issues with children and schools. And funding issues on public projects. The list goes on and on.

I realized it’s not possible to take a comprehensive approach to fixing the food system. Period. You can’t apply any abstract, generalized thinking to this—solutions must be tailored to the ground conditions of each and every micro-environment. And a lot of solutions need to happen everywhere, all at once, in order for good changes to sweep the system.

For instance, one cannot simply say “I will create bottom-up, consumer-level change.” Once you unpack that sentence, you come to the realization that there are all sorts of consumers. Changing their attitudes towards food would accordingly require all sorts of approaches. A busy Wall Street worker is not going to have the same food situation as a mother of five that lives in a food desert. The way you convince them to shop differently, to advocate to their friends, to get involved with the ‘movement,’ is going to vary widely, from the tone you take to the technological channels you employ.

Knowing this is actually kind of liberating. That means that it’s less important to sit around cogitating on the one “best” way to approach this problem. It’s more important to choose an angle, preferably one that you have excellent resources for and access to, and do some real ground work. Tackle it head on, prototype it on real people, build it for real, and release it into the wild.

Now as I really start thinking about what to make for my thesis, I at least know this: I’m going to build it, and it’s going to live out in the world.