Like most, the Pandemic drove me to do more cooking at home over the past 2 years.
There are a lot of benefits to doing this, cost savings and health benefits, but it does come at the sacrifice of time.
Even before you start chopping veggies, you have to answer a series of questions like:
Who all am I cooking for?
Do any of them have dietary restrictions?
How much food should I make?
Will this meal be healthy?
Do I have all the ingredients?
Should I buy this brand of olive oil, or that one?
It can be a lot, and explains why so often fast food can be a luring alternative.
For the purposes of this project, it was less important to interview people who already use Instacart, but rather focusing on individuals who either consider themselves home cooks, or cook meals at home regularly.
I wanted to get inside the heads of these cooks, and find out what they're concerned about when preparing for a meal, and what kind of hurdles they face that make them less likely to cook.
User Interview Questions >
Affinity Mapping >
Interview Takeaways >
- Make meal prep easier for home cooks?
- Make sourcing recipe ingredients easier for individuals with dietary restrictions?
- Give home cooks more control over their meals?
Stepping into the shoes of one of my potential users, I crafted a persona.
This persona, and her family, became a central part of my user testing scenario.
Annie was faced with attempting to use the current app to search for recipes that will accommodate her daughter's dietary needs. This exercise revealed a major stumbling block in planning for mealtime.
Elements of many of these ideas contributed to the final solution.
Taking a quick look at how the market is currently addressing user needs, I created a handful of categories of products that address those.
After looking at a wide range of ideas, I identified adding filters for different dietary preferences and needs as the best way to address the needs of the widest range of user needs.
I looked beyond the world of dietary needs when searching for filtering design patterns to take UI inspiration from. Many filters rely on simple lists with check boxes, others used selectable tags or toggles. I would want to borrow elements from design patterns users are already familiar with, but it'd be nice to make it more engaging, especially given all the filters that would be necessary.
Rigorous research into dietary needs and preferences revealed a very complex world of overlapping restrictions.
This was my initial attempt to bring order to the chaos, dividing restrictions into 3 main categories.
- Ensure the structure of dietary restrictions made sense to people with restrictions.
- Test how well users understood the feature, and how it functions.
- Uncover any usability issues, and inefficiencies in the process.
Testing Questions/Scenario Script >
The prototype I created was built around a scenario, in which participants would take on certain traits, and tailor their dietary profile to match those needs.
- The concept behind the feature was generally well understood, although there was initial confusion about the use case for multiple profiles.
- An attempt to streamline the process of selecting their dietary needs backfired as users felt unsure about their decisions.
- Another issue that needed addressing was click space, with users accidentally hitting click zones and not being able to click others.
Full Testing Takeaways >
After discovering a big sore point during user testing, I recreated the task flow from my testing scenario, this time without the extraneous step.
Extensive Research was conducted on food allergies, and the more I learned the more I understood how complex the Information Architecture needed to be.
Search vs Categories
I was torn between a search based mechanism and a category based solution for users to select their allergies. Categories made more sense when I learned more about how one allergy might effect a whole group of similar foods.
Allergy Pervasiveness
Research showed 90% of food allergies come from just a handful of allergens, so to expedite the process for that large contingent of users, I included a list of common allergens/intolerances that would be the first options when selecting allergies.
Allergen Subcategories
Often people are allergic to entire subcategories so there needs to be a way for them to select all items in the category at once. This was achieved with subcategories. So someone who is allergic to all Tree Nuts does not have to select every type of Tree Nut individually.
Selection Behavior
None of these are exclusionary, if you select one, that doesn’t prevent you from selecting any others.
Selection Behavior
Unlike allergies, each category of dietary choices is exclusionary. Multiple items within the same category cannot be selected at the same time.
Which dietary restrictions?
When choosing what dietary restrictions to include toggles for, I wanted to ensure I was covering the vast majority of important user needs, and not putting users at risk of a diet that would put them in nutritional danger.
Spectrums
Some categories represent a spectrum of options, (ie 'low fat' and 'fat free') indicated by linked options.
With a newly reorganized content structure that adds more refinement to user choices, I explored a number of new design patterns that would accommodate this new structure.
Taking the new content structure into account, I created a new version of the prototype for this feature, focusing this time on tightening the aesthetics to match the Instacart app.
View Full Prototype >
In large households, multiple profile cards may become overwhelming, so I worked on a 'minimized' version that could be toggled with access to the bare essential information and control.
Users are presented with common food allergens, followed by categories of foods that induce allergic reactions. Users can select a whole subcategory such as 'Tree Nuts' or individual allergens.
Many restrictions were mutually exclusive to one another, or existed on a spectrum with other choices, this informed the content structure.
Navigating the world with dietary restrictions
I definitely gained a newfound appreciation for how much more difficult shopping for food and prepping for meals was for people with dietary restrictions or allergies.
Difficult IA Challenge
This was truly a test like I hadn’t faced before when it came to organizing content, there wasn't a 'perfect' solution to be found, and I had to determine what constituted 'good enough' to address the largest number of dietary needs.
Research is a process, not a step
This is a lesson I’m already familiar with, but this was a project where research really had to continue through almost every step in the process. Secondary research had to be conducted on the initially defined problem space, then on comparable solutions to user needs, and finally design patterns for filtering results.
If I were continuing this project it would be important to:
- Conduct User Testing with Hi Fidelity Prototype
- Address any potential problems uncovered in testing
- Get feedback from stakeholders within the company (if I worked at Instacart)
.gif)
.gif)
.gif)
.png)
.jpg){kind=small}
