Detailed business model
Value: User Friendly
Initial UI concept designs
Product database overview
Price list database overview (updated weekly) - maps with product database
Example of entries from the price database (shows which store it's being pulled from)
Example of entries from the product database, including unit weight
UI Shopping List
UI Search Screen
UI Pull out menu
UI Search Screen
UI Shopping List
First draft of our business model canvas
Using a social enterprise model, we’re creating a shopping list app, which will not only save people money while helping them buy healthier food, but will also collect data about their shopping habits, which, when linked with their postcode and household income, will hopefully allow researchers to better understand some of the correlations between eating habits and health outcomes.
We appreciate that one of the big issues with data collection is getting people to regularly and reliability fill in whatever form/survey you’re looking for them to complete - the recent Census is a perfect example of that! What we’re hoping to create is a product that gives users a very clear ‘What’s in it for Me’ by providing an easy to use, intuitive shopping app that can be accessed by the whole family or business, which lets them know how much they can save each time they complete a shopping list, providing a running tally of the amount they’ve saved to date using the app, tracking specials, telling them when they should or shouldn’t buy specific products based on their price history, and offering them tailored sale advice, based on their buying habits.
If we do this properly, by providing users with a free, accessible, intuitive and easy to use app that genuinely saves them money and makes their lives easier, we hope that people will be more likely to use the app regularly enough to provide real, usable data that can then be passed on to UTas and other appropriate research bodies to help inform health research.
We plan to start simple and small, only releasing the app on Android initially, only including products from the two big supermarket chains (Coles and Woolworths), and only providing a fairly limited set of core functionalities in the first instance, including taking a list of products from the user, and returning advice re how much they would save if they chose to shop at Coles or Woolworths, or if they were prepared to split their shop between the two stores. It would tell them how much they’ve saved to date using the app, and show a shopping history. It would track items they wanted to ‘watch’, to see when they came on sale, and would offer advice on when and where they should buy certain items on their list, e.g. we all know that you should never buy Vitamin tablets at full price, as they go for 50% off regularly! Basically, we aim to help people shop the bargains more often than not! As an added benefit, we hope that by saving money on their grocery budget, people will have more 'disposable' money to spend on perceived 'expensive', healthier foods, including fresh fruit and veggies. We hope that the app may also help to break down some of the, often false, perceptions that eating healthy is more expensive.
Looking to the future, we would plan to expand the functionalities to include other shops beyond ‘the big two’, e.g. IGAs and other smaller supermarkets and local producers. We would make the specials more tailored, integrate suggested recipes based on people’s product history, and continue to improve the machine learning element of the project. We would push the app onto IOS, as well as launch a web platform. Once we started to get some credibility and traction with users we may also look at asking some extra questions, like whether you got the flu this winter, and how often you eat out or buy take away food, to try and further increase the value of the data being collected.
In a nutshell, our goal is to create a product that not only improves people’s health, and helps them make ends meet, but also, potentially, will help to inform health research into the future.
You can play with our database here: change the word after the '=' to change the search - use commas to separate words!
Unfortunately, though we invested significant time in building the Android app interface for this project, it was developed in multiple places simultaneously, and at the point at which we tried integrate the different elements, we encountered dependency conflicts, with some libraries strictly requiring different versions of Android - as we'd built this far in, it was too late to start over from scratch. However: we didn't want the major work we'd been putting into the database (backend), to go to waste, which is why we've provided the above link, to let you have a play with our data. The screenshots we've included of the UI are there to give an indication of the look and feel of the app that we would develop further if given the opportunity.
The scraping process was a significant one; working to find the private APIs used by Coles and Woolworths to publish this data was a complicated process in and of itself. We then needed to standardise the data, including weights and measurements. We tracked not only 'specials', but also multi buys, and created 'generic' products to use in the search process. We had to use different processes for each Coles and Woolworths, as they have significantly different set ups. To add to our woes, just as we were settling down to do some significant testing, Coles store went down for maintenance for 5 hours! J #hackathon!
- intuitive, easy to use Android app, free to download and use
- option to donate a percentage of savings back to the project to help with sustainability
- complete product database of Coles and Woolworths stores, updated weekly with price changes
- auto complete when entering products, using elasticsearch with the existing database, as well as suggested items to add based on your search history
- generic 'product' items, allowing people to simply add 'milk' or 'bread' to their list, without specifying brand, size etc. and the App will return the recommended best buy from all the available options (this is then editable, if you don't like the chosen option)
- we will give you advice about where you should shop, based on the items on your list: we will tell you how much you'll save if you shop at Woolworths, at Coles, or if you choose to split your shop between the two (this will give an itemised list of products for each scenario, which you will then be able to edit if you wish)
- increase buying power, by telling users when they should, and shouldn't, be buying specific items, using the historical price data collected from the stores. We'll be able to tell you when specials are actually 'special enough' to warrant adding to your list, and let you know when things are likely to come on sale at a good price.
- shared lists, that are accessible by multiple family members/household members/business colleagues, replacing the combination of sticky notes and frantic test messages currently employed by many shoppers! By having the app on your phone, and accessible by the appropriate group of people, not only will your list be far more complete than before, it will also be more detailed, including brands and weights and varieties, making it much easier to safely deputise the shopping to someone else!
- watch list, which you can add items to until they come up on sale, at which point you'll be prompted to move it onto your shopping list
- tailored advice re specials and savings, based on your buying history
- searchable history, tracking your previous list items, and how much you've saved to date with the app
- a few, simple questions upon sign up that will help to make the data collected more useful to researchers, including ave. household income, number of house members and postcode (if not already collected from your device's location)
- single sign in using your Google account
- expand to incorporate other stores, e.g. IGAs and smaller, local supermarkets
- launch OS app and web interface for better accessibility
- fully integrate machine learning, that will learn your habits, tastes and preferences, and put them into effect in how the app responds to you, and what it suggests and recommends
- integrate suggested recipes based on people’s product history
- network with health organisations to work together to promote the app to our target audiences, and gather feedback on its usability in real world test cases
- work closely with researchers to ensure that the data being collected by the app is relevant to the research being conducted
- once we started to get some credibility and traction with users we may also look at asking some extra questions, like whether you got the flu this winter, and how often you eat out or buy take away food - these questions would be based on advice received from the researchers involved.
Business Model Canvas
A Prezi presentation exploring our Business Model Canvas is available here.
UHack Research Themes
- Better Health: This project aims to collect authentic, real-world data on what food people are buying, linked with their annual household income, number of household members and location. Data on what people eat is famously hard to collect, and though this project doesn't provide information on exactly what individuals are eating, it does provide a different set of data, connecting location, income and food in potentially previously unexplored ways. We anticipate that this data could be used to further research particularly in the areas of policy and prevention.
- Data, Knowledge & Decisions: Not only are we collecting data in new and innovative ways, but we hope that the research conducted using this data may inform evidence-based policy in the health area into the future.