Online, On Paper and In Person: The lightweight methods we used to redesign the mobile UX for local restaurant reviews.

Five simple research and prototyping tools we used to reimagine how locals find the best of what’s around them.

Ideas for our projects in the Lenfest Local Lab typically come from a few sources — journalists, collaborators, the industry and sometimes ourselves. For this app, it was a combination.

We thought about building a local restaurant review product early on but we weren’t capable of it yet. Once we were fully staffed and launched the HERE app — which highlights local news stories about your location — the project was starting to feel within reach. It was around that time people also started asking us:

“Why don’t you build an app for local restaurant reviews?”

Things were starting to fall into place but we also knew we would have to consider a lot more elements of the user experience this time, including people’s food preferences and the best planning features to offer. We also wanted to expand the geographic area of the app out into the suburbs.

A slide explaining where the local restaurant review app fits into our lab’s planning.

We started our research process with this hypothesis in mind: If professional local restaurant reviews were available through a modern mobile user experience, engagement with them would increase and improve.

Because we’re a lab working in the open, we’re sharing notes about how we approached the research and product design process. Here are all the methods we used to guide our work, which we’ll explain below:

  1. Competitive analysis (What’s already out there?)
  2. In-person surveys (What are people actually using?)
  3. Content audit & card sorting (What info do people want the most?)
  4. Online survey (What do people think is most useful?)
  5. Paper prototyping (What features are easy to use?)

1. Competitive Analysis

What’s already out there?

I began my research by looking at other restaurant apps, including products from YelpOpenTableZagatLocalEats, and Chefsfeed. I listed features they had in common, like options to filter results and find restaurants nearby. I also browsed customer reviews for clues about what already was and wasn’t working for people. Then I shared thoughts with the team about how our app could be as useful and also different from what was already out there.

Screens from existing food-finding apps. From the left: YelpOpenTableZagatLocalEats, and Chefsfeed.
  • The main differentiator for our app could be the quality and curation of the restaurant reviews from our collaborators at The Philadelphia Inquirer — offering professional critiques, news, and recommendations by local journalists with knowledge about food, local restaurant culture and the personalities that make up Philly’s food scene.
  • We could also incorporate the most popular features from other apps that make reviews easy to find — including search, filters, and sorting.
  • We also wanted to incorporate some interesting, but less common features, such as saving a list of favorite restaurants and sending alerts.
  • One thing that could be a nice-to-have feature was the ability to make a reservation since so many other apps offer that functionality.

This competitive analysis helped us see and align around the big picture for this app, as well as identify the must-haves vs. the nice-to-have features. We were ready to start talking to people to test our assumptions.

2. In-person interviews

What are people actually using?

After looking at related apps, we wanted to gather impressions from potential users in person. I interviewed 30 people at a nearby coffee shop and local food hall, asking how often they eat out, how they discover where to eat and what apps they used.

Survey Results
  1. Most people go out to eat 2–3 times a week (11 of the 30 respondents). Fewer people eat out 3+ times a week, or just a few times a month. Only one person said they rarely go out to eat.
  2. People like to go out for food and drink mostly for lunch and dinner, and less so for breakfast, a special occasion, happy hour, cocktail and brunch. Not many people go out to eat for work or with colleagues.
  3. When asked about if they like to read restaurant reviews, such as ones written by The Inquirer’s Craig LaBan, almost half of people do.
  4. People find reviews, for new and old restaurants, interesting and think they provide helpful information.
  5. Yelp and OpenTable are the most popular restaurant apps. Some people also go to Philly.com, Philadelphia Magazine, and Eater Philly.
  6. People think getting general information about a restaurant is a must, including the name, address, cuisine type, and phone number. Next, they expect the ability to filter reviews, see nearby suggestions nearby and to make a reservation. The ability to create lists and receive recommendations from journalists or industry professionals weren’t as expected.

You can view the survey here, and read the full analysis here.

3. Content audit & card sorting

What information do people want most?

At this point, we had a sense of the necessary features to include and a better handle on locals’ dining habits. Next, we wanted to look at The Inquirer’s reviews and recommendations and find out what information within them was most useful to people.

Content audit: We looked at a few articles written by Craig LaBan, The Inquirer’s restaurant critic; Michael Klein, The Inquirer’s food and restaurant scene reporter; and Samantha Melamed, an Inquirer features writer who occasionally reviews bars. We were hoping to identify what type of content was shared by all three types of articles.

How we conducted the content audit.

A long list of elements emerged‚ revealing that a lot of information within reviews and recommendation articles could be used in many ways in many new products. We looked at each piece of information as unique and as potentially serving a different purpose, such as photos of the interior or the food, a list of menu highlights, dining notes, price notes or tips about parking.

The content audit results can be viewed here.

Card sorting: From there, I decided to run a quick card sorting exercise to identify which pieces of information might be the most important. I wrote each piece of information down on a sticky note and asked people to rank them — with 1 being the most important information and 4 as the least important.

We didn’t include general information about each restaurant in the exercise, such as the restaurant’s name, address or phone number since the results of our in-person surveys showed us that people always expect to see that information.

How we conducted the card sorting exercise.

Here are the information people ranked, in order, as the most important:

  1. Images of the restaurant
  2. Rating of the restaurant
  3. Menu highlights / what to order
  4. Recommended drinks
  5. Link to the menu
  6. Recommended snacks
  7. Short dining notes
  8. Reservation recommended (y/n)
  9. Restaurant type (Ex. Classic, Modern American)
  10. Average dish price

These results helped us decide what information was important to show people first. For example, each restaurant review has a preview card that shows an image, restaurant rating, cuisine type, and average dish price.

Example of the card layout.

Tapping on the card opens up a page with menu highlights, recommended drinks, short dining notes, and a few non-review items, including a link to the restaurant’s website and a click-to-call button.

Example of a restaurant review page from the app.

4. Online survey

What do Inquirer readers think is most useful?

Before wrapping up the list of features to include in the app, we wanted to ask Inquirer readers directly about what they wanted. We included a survey in The Inquirer’s weekly “Let’s Eat” food newsletter (web version here) asking for feedback.

Our feedback survey link in the “Let’s Eat” newsletter.

We received 50 responses to the five-question survey and the results helped validate what we already were thinking to include in the app. They also helped us decide to add a one-touch “cuisine” filter, saved lists and notification features.

5. Paper prototyping

What features are easy to use and understand?

Throughout the process, we ran many small paper prototype tests to get impressions about icon designs, card layouts, wording, and functionality. In experimentation and discovery work, running small tests to help make small decisions can help things move quickly. By showing different versions of a design to people, watching them interact, and asking them to share their thoughts, we can gather quick directional feedback that helps us prioritize what we build.

Here are a few examples:

Sort and filter features

I showed people two paper versions of the same screen and asked them for feedback. We found that more people preferred simpler options so we decided to use fewer tabs. More notes and analysis can be read here.

List and Map icons

I ran this test to find out whether users would recognize and know how to use the list and map view icons. We found out that the map icon in the second screen was the clearest.

We hope that being transparent about our process is helpful for other teams trying to find lightweight ways to test assumptions with users and move quickly through the steps of product development. That said, each project is unique, and the research methods teams can use should vary as long as they are appropriate for the project.

The Lenfest Local Lab is a multidisciplinary product and user experience innovation team located in Philadelphia and supported by The Lenfest Institute for Journalism.

The Lenfest Institute for Journalism is a non-profit organization whose mission is to develop and support sustainable business models for great local journalism. The Institute was founded in 2016 by entrepreneur H.F. (Gerry) Lenfest with the goal of helping transform the news industry in the digital age to ensure high-quality local journalism remains a cornerstone of democracy.

Thanks to Sarah Schmalbach, André Natta, and Burt Herman. 

Leave a Reply

Your email address will not be published. Required fields are marked *