TGTG is a service that connects users to restaurants and shops with unsold food. Users benefit by buying "surprise bags" from their favourite restaurants and shops at a great discount. And sellers benefit by being able to sell food they would otherwise have to discard. Great discounts and reduced food waste - everybody wins.
Keyword search and map search features are difficult to find in the Too Good To Go app. In usability tests 7 out of 8 participants failed to find the map search tool and abandoned their task.
Too Good To Go is not a delivery app and users must pickup their "surprise bags" at set times. During interviews most participants indicated that a lack of nearby sellers was their biggest obstacle to completing purchases on the app.
The Discover screen on the TGTG app does not currently have any search capabilities. All search tools are located on the Browse screen. During usability tests participants did not realize that the Browse screen provides additional search tools. Unable to find any search tools on the Discover screen participants became frustrated and abandoned their task.
To address this I implemented a suite of 4 search and filtering tools on the discover screen. I also improved navigation and reduced cognitive load by reducing the number of categories that users can see on the Discover screen.
Consumers rarely travel more than 8 minutes to make food purchases. This was one of the findings from a national survey carried out by the marketing firm Access. I conducted a diary study with 3 households and found similar insights - study participants rarely travelled more than 8 minutes to buy food. And when purchasing non-essential food items like snacks and baked goods, participants usually combined these trips with other errands like school runs, weekly food shops, etc.
If users don't have a TGTG seller nearby, the route planner tool allows them to conveniently find sellers en route. Maybe there's a bakery on the way to class. Or a burrito shop on the way home.
The route planner tool allows Too Good To Go to acquire conversions from users who would otherwise find sellers to be too far away.
We want to better understand how users make decisions about where/how they make food purchases so that we can improve conversions in the Too Good To Go app.
Below you can find details for the most important user research methods used for this project. I used the UK Design Council double diamond approach as a methodological framework for the project. There are a few methods not listed here due to space limitations (competitive analysis, personas, journey maps, etc.). If you are interested in reading the unabridged version of this study, see the appendix at the end of this page.
Discover
I conducted an exhaustive heuristic review of the app as is. Evaluated existing interface against list of recognized usability principles.
Approx. 90 minutes
I identified direct and indirect competitors. This included other food waste apps as well as delivery apps and well as a coupon app.
Direct / indirect competitors
5 apps included in analysis
The interviews were structured to better understand how users are engaging with TGTG. Interview participants showed me how they normally use the app and explained their positive and negative experiences.
Semi-structured
Zoom
Approx. 30 minutes
3 participants
Existing users
Walk-through
3 households of varying sizes were asked to keep a record of their consumer purchases over 3 weeks. They kept track of shopping trips, how long they travelled for each trip, and the type of goods that they purchased.
3 households
3-weeks in duration
I used insights from large consumer surveys undertaken by two American marketing firms. The surveys focused on consumer behaviour and the relationship between consumer choice and travel distance.
2021 Valassis "Consumer Intel Report"
2021 Access "Impact of Proximity on Consumer Purchases"
Participants were asked to complete 3 tasks. The tasks were related to search, product categories, and the checkout process. The metrics collected served as a benchmark.
Zoom
Approx. 30 minutes
6 participants
New users
3 tasks
Define
The Valassis Consumer Intel Report described in the surveys section above provided significant demographic insights to create actionable personas.
-
I created a journey map to better understand the user journey with the Too Good To Go app.
-
Develop
For the first prototyping iteration I created mid-fidelity wireframes using Figma.
Wireframes
Figma
The mid-fidelity wireframes were were evaluated using semi-structured user interviews. Interview participants discussed with me whether the UX solution properly addressed their pain points.
In person
Approx. 30 minutes
3 participants
High fidelity prototypes were created using Figma.
Figma
Participants were asked to complete 1 task related to the map search tool.
In person
Approx. 30 minutes
8 participants
Participants were asked to respond to a short survey aimed to measure the users overall experience. The questions asked were related to likely future use and likelihood to recommend to a friend.
8 participants
In this section I discuss the insights gained from the user research conducted for this project. The insights can be broadly divided into two categories: insights about navigation and search, and insights about travel distance. The key insights are listed below. A bit further down I describe the methods I used to arrive at these insights (interviews and usability tests).
1
The vast majority of test participants were unable to find the keyword and map search tools on the Discover screen, the first screen of the app. As a result, most users abandoned their tasks in usability tests.
2
Users were unable to distinguish between the Discover screen and the Browse screen. During user interviews participants were unable to list the differences in functionality between the two screens.
3
Users found the categories within the Discover screen to be useful when deciding what to purchase. However, they found the navigation, search, and filtering behaviour within the Discover categories to be confusing and difficult to use.
I arrived at the insights described above by conducting user interviews with existing and new users of the Too Good To Go app, as well as by analyzing task metrics I collected from usability tests.
I have included below some sample quotes pulled from interview sessions. Interviews took about 20-30 minutes and users were asked to give me a walkthrough demonstrating how they normally use the app.
“I’m not familiar with any of the sellers that I see on the Discover screen. How do I find sellers that I know?”
“There is no reference to location on the Discover screen so I don’t know if I’m looking at sellers that are nearby or far away.”
“I want to find sellers that are right beside me but I can’t figure out how to do it.”
“To be honest, the idea of having to go out of my way to pickup a food bag, when I don’t even know what’s in it, isn’t really worth it to me.”
“I stopped using the app because I didn't have any familiar sellers near me.”
“Some of the categories on the Discover screen don't make a lot of sense to me. In the past I just saved my ”
I used task metrics from usability tests to help me form the insights discussed above. I asked 8 new users of Too Good to Go to perform 3 tasks related to map search, checkout, and use of categories. The tasks and task success rates are below. Most notably, the success rate for finding the map search tool was very low for new and existing users.
Task 1: Use the map search tool to find something for lunch today
1/8
Task 2: Find a surprise bag you like and add it to your cart
7/8
Task 3: Find a vegetarian surprise bag
8/8
Key insights related to travel time are listed below. A bit further down I describe the methods I used to get these insights (surveys and a diary study).
1
Consumers are generally unwilling to travel further than 8 minutes for food purchases. The distance is measured in minutes because this applies whether consumers are walking, traveling by bike, or driving.
2
Consumers generally do not take standalone trips to buy snacks or treats. By and large, trips to purchase snacks or treats are combined with other trips (multi-trips), like errands, or other shopping trips.
3
The number one obstacle for test participants to make purchases on the Too Good To Go app is the distance of sellers. Generally, test participants did not find sellers within a distance they were willing to travel.
I used two large surveys produced by American marketing companies to derive insights related to consumer purchasing behaviour and travel distance. On average, consumers are only willing to travel up to 8 minutes for takeaway. This is cross validates some of the findings from the user interviews, specifically that the biggest obstacle for users is not having sellers nearby.
Percent of urban consumers who travel less than 15 minutes for routine purchases
Average distance that consumers are willing to travel for groceries
Average distance that consumers are willing to travel for takeaway
After travel distance, price is the next most important criteria consumers consider when deciding where to make a purchase
I had 3 households record all of their consumer purchases over a 3-week period. I asked them to record the type of purchase, the name of the shop, how far they travelled to that shop, and whether it was a dedicated trip or if the trip was combined with other tasks.
The findings cross-validated insights from the user interviews and from the surveys. Specifically, all trips for food purchases recorded during the study were of 10 minutes or less. Additionally, I found trips for snacks were never dedicated trips, but were usually combined with other tasks. This is significant for Too Good To Go as a large proportion of sellers sell food items that are not meals (like baked goods, for example.
I transcribed and coded the interview data and did a thematic analysis. Three important themes emerged from the process. They are listed below.
All interview participants indicated that there are not enough sellers near their home.
Most of the interview participants took multiple attempts to find the map search feature. It was a source of frustration as participants scrolled their screens multiple times before finding the search tools.
All interview participants were unable to distinguish the functional difference between the Discover and the Browse screen. Particiants said they would like the ability to filter through the results in the Discover screen categories
Based on the insights above I set up some design goals to guide my design process.
1
Facilitate users who do not have sellers within 8 minutes travel time.
I did this by implementing a route planner. The route planner allows users to plan a trip so that they can stop at TGTG sellers en route.
2
Redesign navigation and search tools, especially on the Discover screen.
I did this by implementing 4 search and filter tools on the Discover screen, which users can see immediately after opening the app. I also made improvements to the search and filtering behaviour of the Discover categories.
3
The number one obstacle for test participants to make purchases on the Too Good To Go app is the distance of sellers. Generally, test participants did not find sellers within a distance they were willing to travel.
The route planner tool displays the route from a starting point to an end point. It also displays the TGTG sellers that can be found along the way, within a distance set by the user. Users are able to filter the results that they see on the map, and they can change their route according to the seller they select.
During interviews users expressed confidence in the familiar interface of the route tool, which is based on the Google Maps interface.
Route Planner
Screen 1
The route planner allow users to set a search distance along their travel route.
Route Planner
Screen 2
A travel route is displayed with a starting and ending point. Sellers near the route are displayed
Route Planner
Screen 3
Users can filter their results so only relevant sellers are visible along their route
Route Planner
Screen 4
When users select a seller on the map the Too Good To Go app suggests a new route and displays an estimated travel time. Users can then use their Google maps for navigation or turn by turn directions.
Route Planner
Screen 1
The route planner allow users to set a search distance along their travel route.
Route Planner
Screen 2
Sellers near the route are displayed and their details are listed in card at the bottom of the screen. User can tap on this card for more information on seller.
Route Planner
Screen 3
Users can filter their results so only relevant sellers are visible along their route
Route Planner
Screen 4
Users can pan the map as they do with other apps like Google Maps. The behaviour of the route planner will be very familiar to users.
Route Planner
Screen 5
When users tap on a seller the app suggests a route that stops at that seller along the way. Tapping a seller on the map brings up that sellers card at the bottom of the screen
Route Planner
Screen 6
Users can tap on more than one seller to compare routes and decide where to go. Tapping on the "Get Directions" link will open users' map app of choice.
I redesigned the Discover screen to include search and filtering tools that are immediately visible to the user. The 4 search tools include a search bar, a map view button, a route planner button, and filtering categories.
During user interviews users preferred the design where the search tools were at the top of the layout. However, I decided to use the third option that leaves two discover categories at the top of the layout because TGTG often uses these categories for featured content.
Discover Screen
Option 1
Discover Screen
Option 2
Discover Screen
Option 3
Subtle changes to the Discover screen make significant improvements in usability.
Discover Screen
Before
Discover Screen
After
I evaluated the high fidelity prototypes through usability tests and user interviews. For the usability tests participants were asked to complete 1 task that could be benchmarked against the metrics obtained during the formative portion of the project.
For task 1, participants were simply asked to use the map search tools to find a surprise bag. The task was marked as success if participants tapped on the map search button. All test participants found the search tools.
I did not revisit tasks 2 or 3 as I did not redesign these portions of the app.
The results from the study suggest that simple changes to the interface, especially with regard to adding search tools to the Discover screen, can produce significant improvements in usability. Task success rates improved significantly by placing the search tools on the front page of the app where they were immediately visible to users who opened the app.
Additionally, adding a spatial component to the Discover categories would provide users with more context so that they can find their favourite shops faster. The advantage of using a map tool in this scenario is that it provides more context to the user without having to provide them with search or sorting tools, which would likely push sponsored content off the screen.
Finally, the study revealed that adding a route planner tool partially addresses the problem of not having enough sellers in some Dublin neighborhoods. A route planner tool allows users to search for surprise bags without having to take the time to make a special trip to collect it.
I have included an appendix for this project to include some of the more nitty gritty details of the process. Click on one of the items below to see additional details.
The personas created for this project were put together from actionable insights derived from large surveys administered by two American marketing companies: Valassis and Access. The demographic insights derived from the national surveys were important and actionable. They are listed below.
