Satisfying the Unsatisfied

Nutritional information is the one thing we’ve heard from our customers all the time over the years, through NPS score, customer visit, member service, account managers, product feedback, etc. But it is like the mission impossible for the company, because we have hundreds of restaurant dishes on the menu for all the locations, restaurants wouldn’t give us their secret recipes; we switch out dishes that don’t get high ratings; it costs a fortune to send all the dishes to the lab and get accurate nutrition facts/label.

If some nutrition proxy can be populated for all the restaurant dishes, customers’ meal selecting experience will be improved and they will select the right dish that fits their diet goal.

Design Process
Identify our target users

I started asking the question around “Do you care about the nutrition?” 99% of people would answer “Yes!”, but do they really look at nutritional facts/labels when they are dining out? With that question in mind, I conducted a qualitative research with 10 users, to figure out what nutrition do they look for exactly and why.

Imaging you walk in Chipotle, and order a steak burrito bowl.

Would you look at the nutrition info?

If yes, what do you look at?

Why do you read those info?

These 10 people fall into 3 groups:


  • Nutri - Pro: who understand the diet impact on health, always look at the nutrition label if available, know exactly how many calories and grams of macros they need daily, can determine what dish to eat by “eyeballing it”, follow a certain diet.


  • Nutri - Conscious: who have a general idea about nutrition label, check the label for the macros, mostly depend on the calories and %DV to determine if the food is good for them. Within this group, some are Nutri-Pro, who have the knowledge but don’t follow any diet; some have limited knowledge about nutrition.


  • Nutri - No: who do not check nutrition labels.


“Healthy means different to different people, it (the Nutritionist’s Pick) needs more explanation for what makes it healthy, eg low carb, high fiber, etc.” - Nutri-Pro


“I would choose a dish based on the nutritionist’s sticker, also look at the dish picture to make sure nothing I don’t like.” - Nutri-Conscious


Different level of nutrition info satisfy different people, based on their knowledge / preference of their diet. Nutri - Pro group can either eyeball the dish and make decisions, or cannot be satisfied until they see the nutrition label; so our target users are the section in Nutri - Conscious group who have limited knowledge about nutrition, but want to eat healthy.

  1. Nutritionist’s Pick - A tag on the dish image with a blurb explaining why it’s healthy.

  2. Dish Highlights - Restaurants label their dishes with tags of “high protein”, “low sugar”, “high fiber”, “low carbs” to indicate the macros of the dish.

  3. Nutrition Ranges - showing the range graph of each category from low to high, based on a 2,000 calorie diet, along with the number range as well.

  4. Nutrition facts for a similar dish - a plugin from myfitnesspal app, just like those who use the app to log their meals, they search a similar dish when there’s no match.

  5. Nutrition Goal - Pre-identified diet goals with ranges of nutrition facts, eg. body builder, a healthy diet, and a controlled diet, then match them to the dishes.


Our product manager sent out a survey to 1000 users to determine which solution we should go with. In order to get unbiased answers, we designed the survey to be not about nutritional information, but about how satisfied are users with the provided information on the menu? And if not satisfied, what can we provide to satisfy the greatest percent of unsatisfied users? In the survey, we identified the unsatisfied/sometimes satisfied users first, then we showed them the mockups based on the complexity, from the nutritionist’s pick to the nutrition goals. Once users hit the answer “yes, I’m satisfied”, we exit them the survey. In the results we received, among the unsatisfied/sometimes satisfied users, most of them exit at the mockup of nutrition range, 52% to be exact. So we got our answer.


After we nailed down the direction, I did another iteration of visual exploration and user testing to see which graph conveys the nutrition range the best.

  • The perception of the gradient color is not matching the meaning of the nutrition, eg. color red means alert, urgency ; but high fiber in our diet is a good thing.

  • A sliding bar on a scale from low to high is very abstract, or seems random to users.

  • A bar chart is harder to read, eyes go up and down instead of left and right, as we are used to.

  • The separated bars makes it not continuous.

Final Design

Empowering Office Admins

Where's My Lunch