I never thought I’d be able to become an entrepreneur. I never thought I’d be able to translate my ideas into reality. Just a year ago, there were all sorts of mental barriers in my head telling me it wasn’t possible for someone so young, and that I was too “busy” with schoolwork anyway. Luckily, I was introduced to entrepreneurship and startups when it suddenly BECAME my schoolwork.
One year holds a lot of information to compile. That's why I'm going to skip over most of it. There's a saying that goes: "20% of the work accounts for 80% of the result." That would probably be the case in my work leading Savior Systems, so I'll focus on the points that showcase my skills and most valuable takeaways that you can learn from.
Before I get into the bulk of this case study, I wanted to mention that I was only able to take advantage of this opportunity to lead other students as entrepreneurs through the Learn to Start program, which had just been brought to my school. I have a short article on my experience up on The Startup Studio's website. This case study will go more in depth on the specific skills I developed in solving unique problems.
On the first day of class, we had to stand up and introduce ourselves to the class with an interesting fact about us. What seemed like a typical icebreaker would prove to be a way to expose just how terrible we were at selling ourselves (in a good way). “I’m not interesting,” “I don’t know,” “I see purple where you see blue but I refused to get it fixed because I liked my world better.” Whoa! I guess there had to be someone who got it right on the first try. Even he would find areas for improvement in the first week. The point of this exercise was to demonstrate the power of storytelling. If you’ve ever listened to Simon Sinek: “They don’t buy what you do, they buy why you do it.” Of course, we didn’t become master storytellers or salespeople right away, or even for the entire year. But everyone in the class made marked improvements. From reading off walls of text on template powerpoints, jittering in place, and hesitating at the first sign of doubt, we would work towards becoming able to predict and answer skepticism before it had to be asked, focus on our best material, and deliver a compelling raison d'etre (See, my French class WAS useful).
Storytelling is an essential element of any business, because your business must have some reason to exist to be trusted and successful in the long term. Why did YOU start this business?
In this example, I would create and execute the idea behind Savior Systems, to tackle a problem that would "save" the world.
There are a couple of different routes you can take to find a problem, as I discovered.
You could look at problems that have already been identified, and search for a solution that has not been fully capitalized on. For Savior Systems, we first focused on broad and far-reaching problems like disinformation and food waste, and then narrowed our focus all the way down to specific niches to test for viability.
You could analyze an existing market and see what common customer complaints exist. Of course, "The price is too high!" is probably not a good reason to start a business, as competing by price is too easy of a leverage point to rely on. We looked at the food industry's major actors, like Publix, Allrecipes, Food Network, and Instacart to see what different customer segments were looking for. While some enjoyed the entertainment and gourmet meals from Food Network, others searched for the most convenient option in delivery services, etc.
The first problem that the group identified and were willing to work with was disinformation. We narrowed down our target audience to teenagers still in school who were present on social media. After some research and deliberation, we presented a solution: a website made to inform teenagers about the issues that matter to them, paired with a strong social media presence, where disinformation is often spread.
There were lots of problems with this, even on the surface. How could we be trusted to determine the validity of news? How would we make money with this, even if we went non-profit? Who would want to use this?
So, we were forced to start over, and for the better. My teammates, despite agreeing to the idea, were not enthused by it, and the product would have flopped immediately, having no paying market or revenue streams.
We found a much greater interest in one of our earlier ideas: tackling food waste with software.
One of the best ways to determine if a market is viable is to see if it already exists. It seemed counter-intuitive to us at first, but having competition is a very good indicator of success. We knew that the food industry was going to exist as long as humans did, because we need food to survive, and food delivery, either from restaurants or stores, has become a convenient trend. If we wanted to create a viable solution, we needed to serve a market that was unsatisfied by current companies.
We weren't planning to miraculously solve food waste and end world hunger with one new fancy app, because software alone can't solve every problem that leads to food waste and hunger. Instead, we narrowed down the problem and looked at where down the pipeline food is wasted. Do farms lose food before handing it to grocery stores? Does food go to waste in trucks? Is food discarded after staying on a shelf for too long? Do consumers waste food after buying it? We found that that last question was the most accurate one, and we were just confirming what we already knew: Food is wasted in homes, and not just with the scraps you leave on your plate (which we ended up not focusing too much on). Oftentimes, people buy foods that they use once and never use again until they expire; people were not making the best use of the fresh foods they had bought, and couldn't keep track of every expiry date in their fridge. We considered this to be a failure of current recipe websites.
To make sure there were actual customers who would want to use an app like this, we analyzed and used websites like Supercook and Allrecipes, which both used algorithms to determine what recipes a user can make with the input ingredients.
Ahh market research... What a wonderful thing. We collected as much data as was relevant about food waste, the food industry, websites, mobile apps, competitors, and trends.
Roughly one third of the food produced in the world for human consumption every year — approximately 1.3 billion tonnes — gets lost or wasted. (Food and Agriculture Organization of the UN)
For richer countries, consumers hold a larger responsibility for food waste. How can we prevent this? What paradigms/habits need to change? “Differing from the situation in developing countries, the behaviour of consumers plays a huge part in industrialized countries.”
For more statistics on the food gap, visit http://ow.ly/rpfMN.
Reducing food waste has become a convenient trend: Actors like universities and stores like Kroger have started movements to prevent and reduce food waste. (Zero Hunger Zero Waste started in late 2017)
We researched every website we thought might be a competitor to our app. Before we invested anything significant into this project, we needed to essentially guarantee it would find an audience and generate revenue. Again, the best way to do this was to look at what the experts were doing. How do Allrecipes, Supercook, and Instacart make money? Who are they selling to?
Analytics websites and companies themselves offer data on their customers, like demographics or bounce rate. We used this data to gauge what kind of customers enjoy Allrecipes and Supercook. We concluded that busy people like parents, usually mothers, use the websites most. Allrecipes and Supercook make money from ads, and Instacart takes a cut out of the delivery fee. Allrecipes also has ads in its magazine as a secondary revenue stream. That is one piece of evidence that showcases another aspect of the food industry: entertainment and celebrities; Food Network takes advantage of this.
Perhaps one of the most important steps in the research phase was to look directly at customer reviews and feedback. From forum posts, app reviews, and LinkedIn pages, we were able to discern a couple key problems that target customers were having: "Allrecipes only lets me input 3 recipes," "Do I have to enter all my ingredients every time I use Supercook?"
While Supercook and Allrecipes were targeting convenience above all else, they still hadn't made their services ultra-convenient enough to keep customers satisfied.
People who use mobile apps for utilities like searching recipes do so because it is fast (perfect for busy people) and can be done while grocery shopping. People who cook are likely in a stable home situation and, especially if they have children, would prefer to stay and cook a quick meal at home. Tools for finding recipes online exist, but for finding recipes that only use ingredients already owned, tools like Supercook are inefficient because all items have to be checked and re-entered whenever the website is used. (This is not only annoying for the customer, but also causes extra food that’s only used in one recipe to go to waste in the pantry or fridge; fresh foods are particularly at risk.)
Now we had identified key issues that customers had with current services, and we could capitalize on those with our product. Still, we had practically no investment and kept the operation as lean as possible. All we needed was a prototype to show to potential customers to further confirm interest. For that, we first focused on our value proposition:
Instantly find recipes tailored to you based on the ingredients you have at home and easily save money on food.
“70% of millennials and 30% of older age groups in the US, for example, use smartphones to find meal ideas, and larger numbers still, to inform grocery shopping”
Eating out is the rich thing to do. If you want to save money and plan for the future, cooking is the trend (at least for the majority of meals). Because the pain point for customers in this segment is lost money, the mere act of visualizing that loss is a huge motivator to reduce food waste and use ingredients efficiently.
Having seen the data on existing consumer habits like mobile app usage, we decided to prototype a free mobile app.
We started with a design that showcased what our vision was for the product. This would change, but at least we could agree on the general idea.
We provided the same instant recipes that Allrecipes and Supercook did, but with much greater individualization and a greater attention to long-term use: You could track your food spending and therefore the food you currently own. Tracking exact amounts and types of food is technically challenging, but we theorized that our counts didn't need to be exact to determine whether a recipe could be made. We did some additional research on how we would best get customers to input their data into the app. Scanning a receipt is the easiest and most straightforward way, if it would work. We discovered that while the receipt could be scanned, only the grocery store would be able to turn the data into food items. Barcodes on individual items show the product, but scanning every item into the app could be tedious. Receipts could be read another way: taking a picture of it and using an algorithm to turn the text into food items. We decided to leave this technical problem for later, and see if the rest of the product could be prototyped.
For individualization, we had some interesting ideas: The app could determine your tastes over the long term and suggest what it thinks may be your favorite foods, much like Amazon's recommended purchase feature. While it was a very cool and modern idea, we knew that customers would want to see the value in the app in the short term as well. A simple way to tailor a program to a customer is to give them a short survey. We didn't want to bore our users before they got to see the full app, so it would have to be only one question: "What cuisine would you like to cook?" The user would select this every time they wanted to search for a recipe.
We could easily narrow down results more than Supercook or Allrecipes, as long as we were able to determine what cuisine every dish in our database belonged to. This seemed simple enough, as many recipes on Allrecipes have the cuisine in the name, like Thai Red Curry or Hungarian Meatballs.
Now comes the landing page users will usually be greeted with upon opening the app. Having selected a cuisine, with the app beginning to infer your tastes, you can scroll through some suggested recipes, or visit the other features in the app.
The sidebar menu is where users can see what's in the “Pantry”, along with some experimental functions like Meal Plan, Connect & Share, and Orders.
We visualize food waste (+ the equivalent money loss) by showing warnings for food that is close to expiration. We then prioritize recipes with those ingredients in our recipe search algorithm.
Our prototype had our value proposition delivered in it, but we also wanted to use our opportunity to engage with customers to test fringe features that, if successful, would be a massive boon to the app.
We added in an internal social media, wanting to test if we could take advantage of the success of Instagram and YouTube food influencers.
We added in a Meal Plan system where a user could select recipes for the week to come, adding to the convenience of shopping and preventing food waste.
We added in a (non-functional) delivery system, assuming we might be able to partner with Instacart or other delivery services.
Now we were ready to talk to customers and collect feedback.
We attended an event for people to come into our school and observe the products students had created. We had a visual prototype that attendees could tap through on an iPad we had at the table as if the app was finished and real. Obviously most features were just for show, but what we were after was the reaction, and if our target customers cared. We were there to gather as much feedback as possible, priming customers to give an honest reaction to the idea of the app. If we influenced their answers in any way or asked leading questions, our experiment would yield false data and we would fail. So we instead talked about the customers' routines and websites or apps they use every day to discern if our app would solve the problems that other services attempt to.
At the end of the day, we had compiled a list of feedback that confirmed our hypotheses: Busy people that currently search recipe websites would use our free app when it was released.
We had successfully identified a target market and prototyped an app for that market. Key problems remained:
Would inserting ads as our primary revenue streams would yield favorable returns?
How can we input data on the individual customers' current foods into the app conveniently?
How can we search for recipes based on foods? (We know Supercook and Allrecipes already do this)
How do users input the foods they already own or recently purchased? There are a couple ideas, though a simple one is by accessing a user’s e-receipts (Available with Publix and Kroger).
Then, the ‘pantry’ could be automatically updated based on new purchases. If a customer doesn’t use e-receipts or needs to enter already existing items, manual input is available, or the barcode for each item can be scanned.
... I could go on and on about iterating, because you could technically do it forever, but you have to launch at some point, so we're aiming for a Minimum Viable Product.
Hope you got something useful out of this summary of the Savior Systems food waste project!