Boosting engagement rate

At Marley Spoon, how the user rate their recipes is not only a question of user engagement, but it is also the ONLY resource for the culinary team to decide which recipes to recreate and which ones to abort. As a new joiner with research background, I took over the responsibility for a smooth product launch and the future feature development.

App design New feature Engagement

Focus: Product design + Usability testing

Date: May 2022

Role: Senior Product designer

Goal

The goal is to allow the user rating recipes more intuitively, to help the user feeling more motivated to rate recipes, and to encourage the user speak out their opinions.

Measuring success – User engagement & Culinary feedback

At Marley Spoon, enhancing the recipe rating feature was pivotal for both user engagement and culinary development. Recognizing that user feedback directly informs our culinary team's decisions on recipe refinement and selection, we aimed to create a more intuitive and motivating rating experience. By decoupling the "done cooking" and "rate recipe" actions, and allowing users to rate recipes at their convenience, we sought to foster a seamless feedback loop. This initiative not only empowers users to share their opinions effortlessly but also ensures that Marley Spoon remains their preferred choice for quality meal experiences.

150%

Strong post-launch growth

After the launch on May 26, Marley Spoon's event count increased from approximately 400 events to a peak of nearly 1,000 events by early June, marking a growth of around 150% in less than two weeks.

10%-15%

Sustained engagement

Following the initial growth spike, Marley Spoon consistently maintained daily event counts between 700 and 1,000, showing a steady engagement level without significant drops (a variation of only about 10-15%).

90%

Stable performance against competition

Despite Dinnerly’s rapid rise in engagement starting in late July, Marley Spoon’s event count remained strong, retaining approximately 80-90% of its peak performance from the initial growth phase.

The design process

01

Research prioritation

How do we plan and implement the most impactful user research methods before launch?

02

Feature prioritation

Which user challenges should we tackle first? After exploring multiple pain points, we chose “rating” as the main focus.

03

Solution exploration

How do we engage product managers and developers from the start to refine ideas quickly? Leveraging Agile sprints allowed us to iterate on UI designs in real time.

04

User flow review

How can an ideation workshop swiftly align the team on the optimal user journey while spotlighting important considerations?

05

Design

How do we craft solutions that balance brand identity, user needs, and feasibility—ready for testing and implementation?

06

Data and feedback

What can both qualitative and quantitative insights teach us post-launch, and how do we use those findings to drive continuous improvements?

Prioritising process of the "rating" feature

Usability testing observing users cooking meals with app in their own kitchen

Key findings - user testing

01

None of the testers naturally rated the recipe.

02

The users didn’t click “done cooking” at all.

03

They are in a hurry to enjoy the meal they just made.

04

Time gap between "cooked a meal" and "rate the recipe".

App store / Playstore review

Key findings - review

01

The user says that they cannot find where to "rate recipes".

02

There are very little amount of "rate recipe" related feedbacks.

Internal user beta testing

Key findings - Self testing

01

Less mentioning "recipe rating" as the internal user testing has its limit. Namely, all internal users are familiar with the existing flow.

02

User finds it annoying that the rating and the comments cannot be changed or edit.

Business input

01

"So from a business perspective dish ratings are the single most important measure for Culinary to understand if the recipe was successful or not, if & what kind of improvement it needs (Dish rating comments), and whether it, therefore, qualifies to be re-run again in a later menu".

02

“We wish to know which recipes our customers like, so that in the future, we could predict and generate recommendations for users.”

HMW workshop & Rapid prototype

A 45 mins “mini-workshop” is being held for the purpose of “kick-off” and “team ideation”. Following the workshop, I quickly put everyone’s idea into low-fi prototypes.

After initial User flow exploration and quick prototype, we split the tasks into different Sprints.

Design iterations

The initial user flow required completing "cooking mode" as a prerequisite for accessing the rating feature.

We “Un-couple” the action of “done cooking” and “rate recipe”. So that the user is able to rate any recipe at any time from the “recipe detail” page with clear instruction.

The 3 options I explored which allow the user to change and edit existing ratings at any time. User can also rate from the “Current” view where the user manage the received box.

Design delivery

User has the total control when when want to cook and rate a recipe

The "pre-loved" and rated recipes stays as a "cookbook" reference to user

Rated recipes are available for each status

Prototype

Design prototype created with Protopie

The results

We launcehd this new feature for Marley Spoon and for Dinnerly in May 26th and July 22nd, respectively. Subsequently, we conducted monitoring of user interaction metrics, including the individual counts of clicks on the 'rate recipe' button. Our analysis reveals a notable increase in user engagement rates following the implementation of the feature.

Since the roll-out, the number of user completed rating significantly increased

Let’s bring your vision to life.

Freelance / Part-time / Remote Full-time

Speake English, Chinese, some Español & Deutsch ;)