This year has been one full of pivots around the world with the pandemic, and our project is no different. One key pivot that was made regarded what our final project idea was going to be for this year. Throughout the ideation phase, one idea that stuck around was some sort of device to either assist the blind community or the deaf community. With this came a bunch of project ideas spanning from "iCanes" to full body suits geared with vibration technology for the blind to special glasses that had pop-up alerts for the deaf. The fact that these ideas stuck around throughout our ideation phase demonstrated that one of them was likely to become our final product idea, and that's what pretty much everyone in the team was expecting as the ideation phase came to a close and it was time to choose our project. However, as more thorough research was done into the ideas and the team became more cognizant of competitors, we struggled to find real points of innovation as time was running out. This was when the initial idea for Sprout came in, and since Sprout presented an innovative, simple project idea that everyone was interested in and hit all the necessary points for this project, we decided to ultimately go with it. However, our ideas relating to blind and deaf help haven't left our heads, and the team may perhaps revisit them next year as we think of another project to flesh out then.
Experiment or Testing Method
A variety of investigations were made throughout the course of making Sprout, including investigating what target market Sprout should be for and what competitors Sprout has, and these investigations are in their respective parts of the profile.
Alpha Testing: In testing the prototype, one challenge that arose was refining the search queries that were used to locate relevant and credible websites relating to the text. One major issue was the returned keywords from the keyword detection algorithm not being relevant to the main topic discussed on the textbook page and thus not returning precise results. By increasing the amount of keywords that were deemed part of a phrase in the keyword filtering algorithm, and in general performing more search queries with phrases from the page, the resulting web pages related more closely to the topics on the webpage.
Beta Testing: Since the app is still mainly in alpha testing, not much beta testing has been performed yet. However, a recording of the "ideal functionality" prototype working can be found as an attachment (same as in the prototype section).
The initial idea for Sprout arose from an idea a team member had regarding scanning in text from literary works and having the app highlight symbols and key literary devices that the author employs. This idea was kept on the backburner during the ideation phase, with not too many team members particularly interested in it. However, as the team approached the end of ideation and was faced with choosing among a few ideas that had not been ruled out, two team members took a weekend to thoroughly look into the book-scanning proposal and see if something could arise from it. From there, the idea for applying the scanning technology to textbooks came to be, and since the team members did not find any significant competitors, it ultimately became a finalist project idea and won out in the end for our product. The key features of Sprout, including generating notes pages and automatic definitions for words, as well as presenting related videos and websites and allowing the user to play with AR-interactive models, were features the team came up with organically with some inspiration (for example, this year's discussion around ZapWorks was where the AR idea came from). Other components of Sprout arose from our research initiatives and discussions with our mentor, such as pricing, what sources Sprout refers to, and how Sprout might actually go about doing its purpose. Discussions surrounding these components can be observed in the research and mentor sections of the profile.
There are two different versions of prototypes for Sprout which showcase different aspects of its progress so far. The end goal for the prototype was to combine optical character recognition, keyword extraction, and automatic Google searching into one device. This proved to be more difficult than anticipated due to the inconsistent nature of the optical character recognition on our device combined with the requirements for the APIs that perform keyword extraction and Google searching. As a result, two prototypes were created - one which showcases the ideal functionality of the Sprout application, and another which demonstrates the attempts to access all three APIs and create a fully functional prototype. Further information can be found in the attached prototype document. Moreover, a video demonstrating the "ideal functionality," hard-coded prototype is attached.
The complete Project Value template is attached.
Another component of our product that can be elucidated here appropriately is pricing and financing for Sprout. This was discussed and explored through primary research as well as team/mentor meetings, and in these sections there are some details, but here we give the complete rundown. The team ultimately chose to go with a subscription model, with having some features free and some restricted behind a $10/month paywall. The free features include presenting related websites/videos, generating non-configurable notes pages, and automatic definitions. The paid features include presenting AR models, configurable notes, practice quizzes, and the ability to view scan history. We believe that this is the best model as it allows users to experience key features of Sprout for free and physically see that it works, enticing them to buy further features. This facilitates better reviews as well (and thus more people using and buying Sprout's subscription). Ultimately, if the company needs more money, we can always rely on selling user data to educational/textbook companies as well. Through these efforts, we believe Sprout will be able to make ByteSized Cookies a handsome profit.