This past weekend I participated in my first hackathon — the first Cleanweb Hackathon in Texas. It took place Sept 21-23 in Duncan Hall at Rice University. Cleanweb is defined as sustainability through software innovation — more specifically, optimizing resource use and accelerating cleantech development. Practically speaking, cleanweb means building applications that tackle energy, waste, water, and other sustainability issues by leveraging web and mobile technologies.
My goals
I had no idea what to expect, and actually signed up two days before the event started. My goal was to 1) meet new people, particularly developers, 2) learn some more programming, and 3) also gain a new experience in trying to ideate and build a minimum viable product within 48 hours.
Since this was my first hackathon, I wanted to get the full all-nighter experience of it, and worked from 9 am Saturday morning till 6:30 am Sunday morning and only slept 3 hours (just so I could function properly and finish building the product).
Our team
I decided to join a team who I, admittedly, thought would win 1st place because they already had 1) a solid, working product in terms of hardware 2) already had a client as Rice University, which signaled validation and demand. The hard part is already done, or so I thought (lessons learned to follow after I provide some more context on our team and product). It was composed of two other Rice students (an applied math major, the other a CS major), a sustainability consultant, an IT consultant, a mobile developer, and me — an econ major with some web dev skills and experience with product management. I wanted to join a team that definitely had technical developers. Ultimately, the CS major, me, and the mobile developer were the ones responsible for actually building our MVP product to demo, while the other three contributed towards pitch development and the business case.
Our product
Our product was called the Revolutionary Trashcan and composed of two parts — 1) hardware wireless scales to measure food waste, and 2) data from those scales that can also be used in data analysis and visualization to save money, increase awareness and induce behavioral change. The actual hardware scales were already built, with the cost of each scale and setup (router, etc) to be around $250 apiece. The original [hardware] product owners already had a contract with Rice to outfit 11 trashcan scales, one per each of Rice’s 11 residential colleges, and several thousand of dollars of funding to ship their products.
The initial intent of the Revolutionary Trashcan is to target universities as clients to help reduce negligent food waste in cafeterias and serveries, and by extension, help educate students to raise awareness of food waste and induce behavioral change.
Here is a video of how the product works (created prior to the hackathon by the other two Rice students, who are the original product owners):
All that was needed was the “cleanweb” component — what do we do with the data, and how do we monetize it?
Our minimum viable product (MVP)
So what did we actually build in those 24 hours? We built an iPhone app (demonstrated through a simulator) that connected to the wireless trashcan scales that listed each of the 11 residential colleges of Rice (i.e., college competition of sustainability to induce behavioral change, more on this later); it was updated in real-time every time someone threw trash away in a trashcan that had one of our wireless scales underneath. We also built an accompanying web app that ideally wanted to have rich data visualization and real-time updates.
My contribution
I basically acted as the UI/UX lead on the front-end work for the web app — including mockups and building out the interface with Bootstrap. I also learned more programming along the way, with the help of the IT consultant, in terms of installing MAMP (Apache, MySQL and PHP on Mac), using PHP to connect to a MySQL database and some basic SQL queries to pull information out of the database.
I also originated the idea to display the data richly and visually through jQuery charts. I would’ve loved to be the lead to connect the data from PHP and MySQL database through a custom API that would translate through JSON and JavaScript/jQuery into a beautiful chart. However, at that point it was already 3 or 4 a.m. on Sunday morning, so the CS major took over that because of time constraints.
One amusing frustration I thought I’d mention — it took me 1.5+ hours and the help of two web devs to figure out why my integration of a jQuery date picker (start/end date) was not working. The solution? Embedding the JavaScript code AFTER the form for the date picker. I had always put it either in the head section, or BEFORE the form for the date picker. Another joy of programming, I suppose.
Our results
We won 3rd place, to my dismay. The 2nd place winning team created a web app to help users discover great places in Houston while using the city’s metro services, and the 1st place winning team created a web app to help calculate and visualize commuter CO2 contributions.
Why? Honestly, I think it was because we did not refine our pitch as well as it could have been — the cost of hardware (which is relatively cheap) was not mentioned or covered, the so-called behavioral change implementation was not well defined, and the market seemed limited to universities and not easily scalable. Our web app also could have been better developed, given more time. If we had more time! If only.
Lessons Learned
1. Importance of the MVP
My initial vision for the web app unfortunately did not come to fruition. Alas, things always take much longer to do (to build, create, etc.) than you think. It came to be Sunday morning already and, according to the law of diminishing returns and productivity, I could not function as well at 2 am to do the actual hard parts, of connecting the data from PHP/MySQL to an API to data visualization through jQuery. Or how I wanted us to have a login system. Or how I wanted us to be able to generate the data into .csv or .xls formats, etc. Or how I wanted us to have a “preview” function. Or how I wanted us to have a “current display” real-time view on the web app.
Because of the ticking clock, I could not accomplish all of that. It was frustrating and annoying to me, and it forced me to concentrate on the bare essentials required to ship the product into a presentable demo that has winning potential.
2. Marketing, marketing, marketing
We could have done a better job of marketing. We could’ve built more business scenarios to think of ways to scale the business and product up. Actually, one of our team members made several phone calls to restaurant owners to see if there might be interest from industry to use our product — and found that they were not necessarily interested in weight measurement of waste, but more so, preventing waste contamination.
Overall, it just boggled my mind somewhat that we had a great, working product, but because of inadequate marketing, we could not sell ourselves as well. It is a great product that we’ve tested and know it will help sustainability initiatives at Rice, but we did not think of additional applications to market it as aggressively as possible.
As Ramit Sethi mentioned in a recent interview on pricing products: If I know my product is good, if I have tested results, it’s my obligation to get it out to the market. I really liked what the interviewer, Patrick McKenzie, had to expand upon what Ramit said:
Patrick notes: I think this is important enough to emphasize, twice. If you got into this business to make peoples’ lives better, and you have produced something which will succeed with that, and you are aware of truth about reality such as “better marketed products beat better engineered products every single bloody time”, then you have an obligation to get better at marketing yourself. To do otherwise is to compromise the value of your offering to the world based on selfish desires such as appeasing your own vanity (“Everyone should realize how great my work is without me needing to tell them”) or indulging your own unspoken fears (“If this were really good, it would sell itself, so if I try selling it, it must not be good.”)
Three of the biggest issues I saw with our pitch was that it didn’t address scalability of the product, data analysis metrics, as well as exactly how it would induce behavioral change in students.
Scalability
Scalability in terms of how to go beyond initial customers of universities (which may be a slow, arduous and bureaucratic process, particularly at public universities) and to the mass market — sports organizations for use at stadiums during high-profile games, restaurants for use to analyze daily waste data, etc.
A key proponent of cleanweb is the scalability aspect of it. We should have done a better job in thinking of possible business cases and customers beyond the scope of universities, which is drastically more limiting.
Data analysis metrics
Our basic metrics are to measure the weight of the waste, timestamp, and added weight (the difference between each consequent weight measurement). One of the product owners simply said, there are infinite possibilities with what we could do with the data — but generalities don’t sell investors. We need concrete details. Like Facebook — with its astronomical amount of data of its 950 million users — but confusion as to how exactly to monetize that data. Data doesn’t matter if you don’t know how to use it or sell it.
We should have done a better job of researching other similar products and analyze the competitive landscape, and do further customer research with restaurants to figure out what kind of data they are looking for, and most importantly, are willing to pay for.
Behavioral change in students
This is a gray beast and involves some deep thinking on the behavioral psychology. It was very intriguing to me, since I am very interested in habit creation and the psychology behind that. There will always be a small group of spirited people interested in being green and reducing waste, but how do you extend that same mentality to the rest of the herd that are not? How do you incentivize them — through money, competition, good feelings, or what? Admittedly, I am one of those people who could care less if I wasted more food or not on any given day. Why? Because it just doesn’t directly affect my life. There is no reward or punishment for throwing away more or less trash. All I care about is that I eat my food and have a place to throw away trash, if needed. However, if there was some sort of rewards program for throwing away less waste that translated into monetary or product rewards, or if some sort of display that I saw on a daily basis of tracking my waste, and so forth — anything that would directly affect my life or bring it more prominently to my attention — that would be a different story.
One of the product owners simply said peer pressure and guilt were sufficient to induce behavioral change, which I considered a gross oversimplification and definitely not specific nor persuasive enough for the judges. What kind of a specific plan do you have to induce behavioral change? Gamification? Competition? If so, how would you design gamification (with rewards, badges, etc) or a college competition to induce such behavioral change? Have you looked at best practices and case studies? How do you know it will work? What is your Plan B if it doesn’t work?
And actually, to combat food waste, the actual battle starts at the beginning — what you put on your plate. How much you put on your plate each time you eat a meal will also affect how much food you waste. A way to exploit this to reduce waste is control portion sizes, obviously, but at a higher psychological level, to persuade customers to eat less to be more fit, look sexier naked, feel healthier, etc. People care about those subjective things much more than objectively helping the environment.
We could have done a better job of constructing a specific plan to address the behavioral change argument, since it is part of our value proposition of educational outreach.
3. Networking
I think I could have done a better job of networking and meeting new people. I guess I was mostly focused on fleshing out the web interface, discussing the business case and pitch and ultimately, just building the product to ship. There were some people from industry present throughout the hackathon, including representatives from Houston’s METRO system, Waste Management and more. In addition, I’d say 70% of the hackathon participants were actually working professionals, and the other 30% were perhaps students. In retrospect, I should have spent more time to also meet other people from other teams.
In other related news
The three most environmentally progressive U.S. presidents are Theodore Roosevelt, Richard M. Nixon, and Jimmy Carter. Another instance of how “cleantech” today is expensive and slow, as the wind power industry is withering because of tax credits — further reinforcing the need for “cleanweb” solutions (in my opinion). Speaking of trash waste reduction with the Revolutionary Trashcan, Dallas is challenging its big oil, big cars and big sprawl to redirect 84% of the trash that currently heads to landfills through its Zero Waste plan.