HelloFresh Design Sprint
to improve efficiency of the recipe development process
HelloFresh- The largest weekly recipe kit delivery service in the United States
Shannon Wackett, Dio Wong, Andrea Lau, Albert Lee, Naomi Mackeown, Myself
Figma, Zoom, Pen + Paper, Slack, InVision
Asset Collection, Prototyping
Users + Audience
The Hellofresh recipe development team
Time: limited to 5 days
Users: didn't have access to recipe development team for interviews/testing
I used empathy and effective communication to help ideate and design a solution to make HelloFresh's recipe ideation process more efficient.
Founded in 2011, Hellofresh has quickly risen to become one of the biggest meal kit delivery companies in the world. They rose from 250,000 users in 2015 to 1.3 million in the third quarter of 2017. The pandemic brought an increased demand for their product, resulting in a massive increase to 7.3 million users worldwide by the first quarter of 2021.
The current recipe development system has struggled to scale alongside this growth. The current system lacks efficiency, uses too many different software programs, and doesn't have a high enough recipe output to meet company expectations and customer demand.
The current recipe development process.
As a growing company, HelloFresh is looking to increase their recipe output in order to give customers more options and to cater to more dietary needs such as gluten-free diets, vegetarian/vegan diets and specific allergies. Their current process for recipe development isn't as efficient as they would like, and they are looking to increase recipe output without hiring more employees.
Increase Recipe Output to Meet Demand
1. Discover user pains, motivations + behaviours
2. Increase efficiency during recipe development
3. Create a platform for increasing recipe output to match business needs
I worked with 5 other designers to ideate a solution from concept to completion. I was the designated Decider in the group and played a major role in ideation, asset collection, prototyping and presenting the final idea to stakeholders.
Hellofresh's brief revealed that their recipe development could be improved upon. To better understand the recipe development process we conducted an interview with a recipe developer from an outside company.
The Food Dudes Restaurant Group
Our interview revealed that recipe developers find the ideation process the most time-consuming, and that they sometimes struggle coming up with new, innovative recipe ideas.
Our interviewee indicated that they typically start with 1-2 ingredients and then use outside sources, such as Serious Eats or Great British Chefs to help with inspiration.
They also take highly rated dishes from past menus and make iterations based on seasonality or dietary concerns.
Based on themes from our interview, we focused on the ideation stage of the recipe development process. Inspiration doesn't always come easily, so we really wanted to focus on a solution that accelerated that process, as well as incorporating previous customer feedback right into the software.
We prototyped an internal software program which allows recipe developers to upload their recipe brief and use AI technology to generate food pairings. The system would also automatically pull up the highest-rated past recipes so they could quickly iterate and produce a brand new recipe based on seasonal ingredients and/or dietary specifications.
Target Selection: We identified the ideation stage as the most critical moment of the user's experience
Software landing page- recipe briefs are uploaded and developers can leave notes for other colleagues
Recipe card for users to auto-generate food pairings and pull up recipes based on customer insights.
In the short amount of time we were given, we achieved our goal of prototyping a product to help speed up the ideation process.
We found that while using AI to generate ingredients helps save time, we don't think AI could completely replace the recipe developers (yet). Recipe developers would still have the work of choosing the ingredients they wanted based on the recommendations, and would still need to ensure there was balance in the recipe (ie. fat, salt, acid).
Moving forward, we would need to do more usability testing to determine that this is a product that could truly ease the pain points of ideation for recipe developers.
Further developing the AI in order to make recommendations based on seasonality and popular trends could further reduce the time needed during ideation, as the recipe developer would just need to make micro-iterations to the flavour profiles.
Building out the software to include a communication portal between developers and testers would further improve the overall process and eliminate some of the extra steps that exist in the current process.