Openfit launched in February of 2019 as Beachbody’s new direct-to-consumer product. Like Beachbody on Demand, it offered streaming workouts, however previous market and user research made clear users wanted a holistic health product. This meant the addition of simple, effective nutrition solutions. This project aimed to increase the value of the product and retention by:
- Determining what kinds of nutrition features both appealed to users and actually delivered the results users wanted
- Shaping healthy choices by providing insights into users’ eating behaviors
- Delivering solutions that were not just elegant and easy to use, but actually dynamic and engaging enough to become a longterm lifestyle
- Creating datapoints that could be leveraged by possible future gamification
- Congratulating users on their accomplishments
Sneak peek of a few final screens
There were many already moving parts when we started this project. Openfit had just acquired Gixo, a company offering live, trainer-led workouts. We were in the midst of integrating the Gixo functionality. This created constraints on engineering resources.
There was a hard deadline. We needed to launch with enough time for a v2 before the end of the year. (Not surprisingly, January is a big deal in the health and fitness market.)
While nutritional offerings were new for Openfit, they were not for Beachbody. This meant our team had an established relationship with the team responsible for nutrition content creation. This team was science-minded and knew how to get people results. We knew from the outset what data points this team wanted our programs to be shaped around, namely caloric intake and macros.
It also meant we had an idea of how receptive the Beachbody audience is to nutritional products. While the Openfit audience skewed slightly more female, the Venn diagrams were close to a perfect circle. Therefore we loosely used the Beachbody numbers to help create our KPIs.
We started with some competitive analyses. In addition to using the apps ourselves, we ran tests on UserTesting.com, asking prospective users to try these competitors and give us their thoughts. We also asked what nutrition apps, programs and diets they tried in the past and how much success they found.
From what we learned, market leaders generally fell into three camps:
- Apps that tracked what users ate (MyFitnessPal, Lose It, Plated, Yazio, Lifesum)
- Apps that told users what to eat (8fit, BetterMe, Centr)
- Apps that told users how to think (Noom, Rise)
For apps that focused on tracking, the food database was key. MyFitnessPal was the household name for a reason: they have the largest database. Users loved that they could eat almost anywhere and find it on MFP. Other than that, the offerings were largely the same across these tracking apps.
For apps that offered specific meal plans, customization was important. BetterMe did a great job of onboarding users in a way that built excitement. The questions were pertinent and specific. The user knew this program would be truly customized.
Screenshots from BetterMe’s nutrition onboarding
Equally important was flexibility. Even the best customization will serve up meals the user just isn’t into at that moment. Both BetterMe and Centr provided ways of switching out unwanted meals for something nutritionally comparable.
Screenshots highlighting BetterMe’s and Centr’s UI for switching meals
There were very few apps in the emotional support, “live nutritionist in your pocket” vein like Noom and Rise. It was difficult to find a significant number of users that had direct experience. However the two people we surveyed who used them (one Noom, the other Rise) said they were effective, long-term ways of approaching nutrition. It was just unfortunate the cost was so formidable.
When asked why something didn’t work or why they stopped using an app, users tended to give one of the following answers:
- Too difficult/strict to follow
- Not enough time (mostly for things like meal prepping)
- It was too tedious; they got bored
- The app was too expensive
Tracking worked for many, however it was not sustainable. It quickly became tedious. Users would assume they could keep track in their head, and soon they were off the wagon.
Prescribed meal plans worked for many, however meal prep was time consuming. We also saw a drop-off after a few weeks, especially if there was a lack of flexibility.
Nutritionist-supported apps seemed to do the most for shaping longterm lifestyle-shaping changes, however the overhead was high. Therefore the cost to the user was high. While this angle had potential, it was clear we didn’t have enough time or resources to pursue this route in v1.
It became clear nutrition was a very personal and unique journey for each user. Therefore we needed features that were:
- smart enough to meet a user where they were on said journey, whether that be simple tracking or full meal plans
- customizable for different dietary preferences and restrictions
- flexible enough to allow users to swap meals, eat out occasionally, generally not be perfect
Because of that looming deadline, we had to get these features somewhat out-of-the-box. As luck would have it, Beachbody had an existing relationship with a while label service called Nutrio that provided meal tracking, meal plans, water tracking, etc. It was time to test if Nutrio was the right partner.
Using Nutrio’s demo, we ran 3 sets of tests on UserTesting.com. We started with feature set tests to ensure Nutrio checked the boxes established during research.
We learned prescriptive meal plans, browsing recipes, dining out guides, weight tracking and meal plan-generated shopping lists had the most traction. Users were somewhat split on water tracking. Most importantly, the vast majority of users said this was a product they could see helping them and they would actually pay for it. Deck available below:
Then we dug deeper on those features in usability tests. We knew we were only using Nutrio on the backend, however we still wanted to narrow in on glaring issues within their features. Among those found were:
Users found the dashboard both overwhelming and directionless.
Also, some users were put off by the focus on calories.
The process for generating a shopping list was far from intuitive.
With stakeholder and engineering sign off, we felt confident moving forward with Nutrio and into the concepting and prototyping phase.
We had our desired feature set, however it was clear not all features would make it into v1. We had to make some tough choices. After conversations with engineering, it was decided weight tracking and the shopping list would have to be fast follows. Also on the roadmap, just a little further down the road, were:
- giving users encouraging progress updates at meaningful times
- intelligently moving from basic tracking to meal plan tracking and vice versa
- reengagement and disengagement-prevention initiatives that reached out to users at key intervals
- partnering with a food delivery service to have ingredients delivered to users’ doorstep
With those sacrifices in mind, we set out to do some rapid prototyping and testing. Due to external forces, we had to move into higher fidelity mocks earlier than we’d like, so even these early prototypes are pretty high fidelity.
We started by stitching together a complete flow in InVision and doing a hybrid market fit / usability test. We asked users to complete specific tasks and asked if they could see a use for this product.
The response was overwhelmingly positive. The market fit was there, however a few areas needed definite improvements. Among those:
The page served after a user completes onboarding (what we called the summary page) needed to be more motivational and congratulatory.
The recipe detail page was confusing. Combining tracking functionality (updating serving size) with meal prep options was a lot to process.
Recipe Detail Page
How much functionality is too much functionality?
We did a few more passes at the recipe detail page. On some we kept functionality the same but visually minimized it. This helped, but still muddied the waters, especially for stakeholders. In the end, moving the ability to update serving size to the dashboard and keeping recipe details dedicated just to the recipe was the way forward.
The final design for the recipe detail page. A user can swipe left to right on the hero/title card to quickly browse recipes.
Infusing emotional design and clarity
Users were confused by the pie chart. They didn’t understand why a chart would show percentages while the accompanying numbers were in grams.
The final design can be seen here.
Conversations with the team responsible for creating the individual macro and caloric thresholds lead us to ranges as opposed to one-number limits.
We also moved toward an animated header with confetti and celebratory messaging to emphasize the enthusiasm for this commitment and journey.
Balancing visual appeal and usability
While the dashboard interface was well received, it was important enough to still rev on. We ran a handful of A/B/C user tests showing different visual treatments and slight ux variants. Here are just a few:
The last option ended up winning. Users liked the clarity of the serving size and Log, Remove and Replace buttons.
A few more revs were taken on onboarding, but they were largely visual. Major updates and additions would have to wait.
Wait, what about delivery? Since we were operating in an agile environment, delivery was continual. Developers and designers attended every stand-up. Nothing was designed that wasn’t possible.
Openfit Nutrition launched back in 2019, but I moved on to a new company before its launch, so unfortunately I never received performance metrics.
Since then Beachbody has sunset Openfit. It was a good run, pun fully intended.