UX CASE STUDY
Reimagining the new user signup experience
Cookpad is a tech company with a goal of building a community platform for people to share recipe ideas and cooking tips. By solving the problems related to everyday cooking and encouraging more people to cook, Cookpad believes that cooking is the key to a happier and healthier life for people, communities and the planet.
- Started out in Japan in 1997, headquartered in Tokyo, Japan and Bristol, UK for global operations.
- World’s largest recipe sharing community, present in over 74 countries in over 32 languages.
- Nearly 2 million Premium Service subscribers in Japan.
- Over 6.7 million recipes created by active users on the platform.
- Over 96 million monthly active users (MAUs) via desktop and mobile properties.
I was the sole Product Design er of this project. I also receive constructive feedback from two fellow designers in order to improve and get other perspectives about what I was designing. I collaborated with a Product Manager, UX Researchers, Product Marketing, and Development Team to research, scope, plan and execute.
This project was designed, developed and executed in early 2019, and I did not have control over the next iterations of the product.
1) To evaluate the existing Cookpad signup experience, and overhaul it to provide users with a simple and delightful path-to-signup and membership experience – one that truly makes them say ‘cooking is fun‘, a tagline synonymous with Cookpad’s mission.
2) Increase website newly registered users (NRU) in Q1 2019.
I followed a modified design process where the initial requirement gathering and research was based on evaluating the existing signup experience and taking it forward from there.
Together with the data team, the first step was to understand how Cookpad functions as a business, metrics around user’s usage behavior and the implications on the business goals.
Cookpad’s business model is freemium and monetizes from Premium Service memberships.
The nature of content on the platform is user-created rather than businesses/brands.
Search had become the primary method of discovery in finding cooking inspiration.
- User engagement is essential for enhanced and personalized user experience.
- In Cookpad’s biggest markets like Japan, Indonesia, and Spain, Cookpad.com appears on the first page and ranks #1 of Google for most popular keywords.
- Over 95% of total monthly active users (MAUs) was via web, and 74% of them was via search engines.
- Over 90% of traffic on the web were guest visits and signup rate is at a low 0.04%.
To gather qualitative data, our UX researchers conducted home interviews with 10 people who frequently looks for recipes online but are not Cookpad users.
These are acquisition related insights:
I looked into other user-generated content platforms and news sites with similar freemium model as Cookpad to see how they are converting visitors to signup. Medium and Tech in Asia gives a limit of 5 articles for free members and visitors. Pinterest asks visitors to login after scrolling down several pins in a board, while Quora prompts a signup modal when a visitor tries to go to another page.
The next step was to evaluate the existing user flow of recipe search.
Google search → Cookpad search result → Recipe page → Signup modal
This gradual engagement process has resulted in a better UX by demonstrating the app’s value before asking to signup, but the signup rate is still considerably low at 0.04%. We saw this as an opportunity to improve the existing signup flow.
Ideating a solution
Hypothesis: Visitors are given no reason to signup and tend to treat the platform like any other recipe site that is owned by media which has a different business goal.
These are some HMW questions we came up after an ideation session:
We settled on this HMW question:
HMW ask visitors to signup at a specific stage of content access
When to ask visitors to signup?
To answer this question we brainstormed the following hypotheses:
- If too soon, visitors may get annoyed from not getting the value of our platform – hence, don’t want to signup
- If too late, users already get what they’re looking for and don’t feel the need to signup
We inquired our data analysts to visualize the number of users grouped by the number of recipes they viewed in a single session. From this chart, it was determined that ~10% (2%+2%+1%+1%+3%) of our daily visitors read more than 5 recipes.
For this experiment, we decided to place an intercept after recipe page #5 to target those 10% visitors.
Percentage of user views / number of recipe pages
Given the short time to launch the experiment, we decided to proceed with a simpler design as MVP. For this experiment, we wanted to try a modal that doesn’t have a (x) close/dismiss button, therefore, users are required to signup in order to continue to browse more content. We took this path to see how extreme the results of this experiment could be.
We launched this experiment on the Indonesian site so we could launch faster without having to deal with timezone differences and involve content writers. We also had several copywriting ideas so we conducted multivariate testing scheduled to launch on the next day.
Multivariate testing using 4 different copywriting in Indonesian
After receiving results from the initial experiment and gathered feedback from the broader team, I continued exploring more iteration ideas. Even though the results are mostly positive, we noticed an increase in failed and exit rate on signup. We wanted to focus on making the signup experience better this time.
Explored several signup modal ideas and count limit interface
Expandable/collapsible count limit tray
Adding a recipe count limit has potential to create a good transition without abruptly asking the user to signup.
Showing the benefits
Communicating the benefits help visitors understand better about Cookpad and give them a reason to be part of the community.
Web desktop view
Improved login page
Login buttons are moved from tray (in v1) to modal considering the majority of Cookpad’s web users have limited screen real estate at 360*640 pixels.
Button placement is moved to the bottom for better reachability. Previously separated ‘Login’ and ‘Register with email’ buttons are merged into a single button.
Adding this modal is a good way to welcome and onboard new users. This may be used in upcoming iterations for feature introductions.
We could see great results coming through this experiment and positively impacting the company as a whole. In just a few days, we saw a massive jump in our NRU and user actions in the platform.
increase in daily registrations
increase in recipe page visits
increase in recipe search
We concluded that adding an intercept was effective to convert more visitors to signup and the increased newly registered users have also grown user stickiness. This was a sufficient signal to get buy-in from other regions and continue with a global rollout in the following weeks.
Build, Test, Repeat: Testing ideas as soon as possible provides data-driven business decisions about what to do next and minimizes the risk and cost of creating features or services that don’t bring impact.
MUCH MORE TO BE DONE
While the results was mostly positive, the product is far from perfect. I plan to continue the iterative cycle of testing and updating the prototype. I will use Zeplin to hand off the design to the developers when I have made all improvements to the prototype.