Sore Thumbs and Empty Stomachs: Design for Decision-Making

A UX Study on the Postmates iOS app experience


It’s Friday night, the lights are dim, and the mood is set. Finally, some quality time with my true loves: my bed and my computer.

Am I in the mood for Friends tonight? Or do I want to watch The Office?

I continue scrolling through Netflix for the next 30 minutes before finally making a decision.

Hmm yes, I do want to watch the Halloween episode of The Office Season 9 for the umpteenth time. Or…do I?

I often spend more time browsing Netflix than watching, and more time thumbing through Yelp than eating. Each day, I make dozens of choices with my thumb. Sometimes those choices are easy, sometimes they’re not. And when they’re not, they stick out like a sore thumb and frustrate me.

Looking back at 2016

We all want to make good decisions that maximize happiness. And let’s be real — that’s not always easy.

On-demand technology has brought almost anything to our fingertips. But with it, overwhelming possibilities. Let’s use food delivery as an example. There’s the convenience of getting food delivered, but also the inconvenience of deciding between 300+ restaurants. I’m almost happier when I only have the 5 restaurants down the street to choose from.

So I ask myself: what is on-demand/time-saving technology good for if it paralyzes users with layers of decisions? This isn’t the way it has to be.

We’ll put the spotlight on a popular on-demand service: Postmates.

Why You Should Care About Postmates

Postmates isn’t any ol’ food delivery app. It uses its network of urban couriers to move goods around cities, delivering “lunch, dinner, groceries, office supplies, or anything else” to anyone. While Postmates is most popular for food delivery, they’ve been steadily introducing and pushing more types of goods to users (e.g. clothing, gifts).

As with any business, as Postmates scales the types and number of offerings on its platform, they’ll want to consider how to organize it all and help users make decisions.

So, I challenged myself to think of how the Postmates user experience can be improved. And I had a hunch that it had something to do with one of my favorite things to do — making decisions.


My Process

I used IDEO’s Design Thinking process to structure my problem solving:


STEP 1: EMPATHIZE

The first step of any effort to improve user experience is to talk to users and observe them interact with your product. So, I did just that.

Assumptions
1) Most users use Postmates for food delivery 
2) Postmates wants to further develop and promote deliveries of goods beyond food

Guerrilla Usability Testing

Assuming most users use Postmates for food delivery, I poached unsuspecting lunch-goers in the Financial District and Yerba Buena Metreon (food court) areas of San Francisco to ensure that users were in the right mindset (hungry). I asked users a few questions about their habits and behaviors, and had users complete three basic tasks using the Postmates app:

  1. Choose a restaurant you would order from
  2. Add an item to your order
  3. Edit your order

STEP 2: DEFINE

After testing with 7 users (why just 5 users reveals valuable insight), I organized my findings with sticky notes. Each color represents a user, and each sticky note represents a pain point or observation.

Insights by User

Next, to identify common pain points, I grouped my observations into themes.

Affinity Mapping: Grouping by Themes

Prioritizing Needs

To identify the most salient problems, I evaluated each pain point along two dimensions: Importance to Users & Importance to Postmates. I based my evaluations on frequency, feasibility, and magnitude of impact.

This process revealed that the biggest points of friction occurred when users were deciding which merchant to order with. Once users made a decision, the rest of the ordering process was pretty smooth sailing. So although task completion was high, frustration was equally high.

Specifically, frustration stemmed from issues with two fundamental steps of effective decision-making: 1) identifying options, and 2) weighing the evidence. And at the root of these issues, respectively, is 1) a confusing information architecture, and 2) a lack of information and functionality to efficiently analyze options.

Core Problem A: Hard to identify options because of confusing information architecture

  • Pain Point #1: Users didn’t know what Postmates Plus was
  • Pain Point #2: Unclear distinction between a curated list of merchants and an individual merchant

Core Problem B: Hard to weigh evidence because of a lack of information and functionality to efficiently analyze options

  • Pain Point #3: Not enough merchant information upfront/available
  • Pain Point #4: Unable to sort, and thus compare options efficiently
  • Pain Point: #5: Merchant photos are not helpful or clear

I annotated my findings below:

Personas

Based on my research, I developed two personas that would help inform my designs. The key difference between the two types of users is their decision-making behavior. Some users have an idea of what they would like to order, while others have absolutely no clue.

Given my initial assumptions, these personas are based on users who use the app for food delivery. That being said, I do consider Postmates’s goal to expand beyond food merchants in my designs, later on.

STEP 3: IDEATE

After talking to users and understanding their needs, I was ready to think of solutions.

Objective
How might we help users make deliberate, thoughtful decisions in an efficient manner that will maximize satisfaction?

Hypothesis
Redesigning the app (as below) will help users find and make decisions in a more efficient and less confusing manner.

Potential Solutions

Core Problem A: Hard to identify options because of confusing information architecture

  • Pain Point #1: Users didn’t know what Postmates Plus was
“Postmates Plus…does that mean I have to pay more?” 
“Is it special locations or faster delivery?”

4 out of 7 users were confused by the concept of Postmates Plus (flat rate $3.99 delivery fee for select merchants). Users hesitated to select a Plus merchant because they thought “Plus” signified a premium service, meaning they would have to pay more. The reality is the opposite: Postmates Plus offers the cheapest delivery option.

Solution: Rename ‘Postmates Plus’ to ‘$3.99 Favorites’, which is more reflective of what the feature is. Also, clarify and make header text (describing what the feature is) more prominent.

  • Pain Point #2: Unclear distinctions between a curated list of merchants and an individual merchant.
“Why do they put lists of restaurants in the same section as restaurants? That tripped me up”

Solution: Create ‘Explore’ tab/feature, one dedicated to lists of merchants and can serve as a centralized source of inspiration when users don’t know what to order. Remove ‘Free’ tab and turn into a pop-up on ‘$3.99 Favorites’ (formerly ‘Postmates Plus’), so that controls on the tab bar are consistent in function.

Below are my proposed changes to the app’s information architecture to make it more intuitive:

Core Problem B: Hard to weigh evidence because of a lack of information and functionality to efficiently analyze options

  • Pain Point #3: Not enough merchant information upfront/available

Users base their decisions on delivery time, price level, and Yelp ratings, and want to know that information upfront. Postmates currently only displays estimated delivery time and type of good. One user even exited the Postmates app to open the Yelp app.

Solution: Provide price information and Yelp rating. Realistically, pulling ratings from the Yelp API may not be feasible due to competition with Yelp’s Yelp24 food delivery service. But if Postmates were to strike a deal somehow, in-app Yelp ratings would enhance the user experience tremendously (seriously).

  • Pain Point #4: Unable to sort, and thus compare options efficiently

6 out of 7 users were frustrated by having to scroll through all the merchants. Users wanted to sort by delivery time, price, and Yelp rating (if displayed). Although sorting by rating may not apply to non-restaurant merchants, the improvement in user experience from the change would trump this minor incongruence.

Solution: Add Sort function to allow sorting by delivery time, price and (Yelp) rating. This will help both types of users (decisive, indecisive) streamline their selection process.

  • Pain Point: #5: Merchant photos are not helpful or clear
“These pizzas look like they could have come from the same place. I don’t get a real sense of what I’m ordering from this picture”

Merchant photos typically only display a single item and have a dark overlay, which prevents users from drawing clear conclusions about a merchant. Further, it’s difficult for users to quickly scan the screen for key decision-making information (i.e. delivery time) when the text is over a photo and center-aligned.

Solution: Replace merchant photos with aerial-view photos displaying multiple items. Remove dark overlay. Left-align and move info text below photos.

Task Flows

To visualize the potential steps users could take and the impact of my proposed solutions, I created a task flow of the current app and another with my solutions implemented.

STEP 4: PROTOTYPING

Lo-Fi UI Sketches

With this improved task flow in mind, I began sketching wireframes of my solutions.

Hi-Fi Prototype

To test my hypothesis, I prototyped my potential solutions below. When applicable, I labeled my annotations with the corresponding pain point it addresses (listed again):

  • Pain Point #1: Unclear what Postmates Plus is
  • Pain Point #2: Unclear distinctions between a curated list or merchants and an individual merchants.
  • Pain Point #3: Not enough merchant information upfront/available, and difficult to read
  • Pain Point #4: No sort functionality. Unable to compare options in an efficient manner
  • Pain Point #5: Photos not robust enough to be helpful

Check out my clickable prototype here!

STEP 5: VALIDATE

Test

To validate my redesigns, I asked 5 users to complete tasks based on users’ key decision-making factors (delivery fee, delivery time, price, rating). I also did a comprehension test on the controls on the bottom tab bar (Anywhere, $3.99 Favorites, Explore). My findings:

  • Although 2 out of 5 users were unsure of the $3.99 Favorites and Explore tabs at first glance, they understood once they tapped into the tab and read the header text. This is an improvement on 4 out of 7 users who were confused by the original Postmates Plus, even after tapping into the tab.
  • All users used the new sort function when prompted with scenarios, and found it helpful in expediting the selection process. E.g. “Say you want to save money, what would you do?” User navigates to $3.99 Favorites tab and sorts by price.

Iterate

To alleviate the slight confusion around the Explore tab, I changed the icon from a file cabinet to a star, and renamed it to be more reflective of the content it contains: “Featured” merchants. The tab now resembles what users are familiar with when they browse the Apple app store.

Small iteration on my tab bar redesign

Overall, with my changes, users were less frustrated and spent less time finding a merchant that met their decision criteria.


Conclusion

TLDR; Making decisions is tough. Decision-making within Postmates can be made easier by simplifying the information architecture and making information easily accessible to analyze.

Conducting this case study was a great opportunity for me to explore tackling issues of scale. As companies grow their offerings, they’ll want to help users make decisions amongst the options. Otherwise, users won’t make a choice at all.

This was also an opportunity to flex my decision-making skills (since design is making deliberate choices). Making choices is hard enough, so it was interesting essentially decision-making for decision-making.

Anyway, thanks for reading! I hope you enjoyed this post and would love to hear your feedback. If you want to chat about design, cats, or anything else — feel free to reach out via email at leung.leanna@gmail.com, on LinkedIn, or in the comments below.

Disclaimer: I am not affiliated with Postmates.

Thanks to Daniel Li and Geronimo Ramos!