How can we help users onboard their community to our app?

A case study of when I designed a feature from scratch, growing Cookpad's DAU by 8% in Taiwan and Spain.

A case study of when I designed a feature from scratch, growing Cookpad's DAU by 8% in Taiwan and Spain.

A case study of when I designed a feature from scratch, growing Cookpad's DAU by 8% in Taiwan and Spain.

Company

Cookpad

Role

Lead designer

Area

Mobile Product Design

Project showcase laptop mockup
Project showcase laptop mockup
Project showcase laptop mockup

Background

Background

Background

As a recipe sharing platform, most of Cookpad’s users are home cooks who love to share recipes in their own social groups. If only they each would invite their group members to Cookpad!

What do they share in groups?

What do they share in groups?

What do they share in groups?

I interviewed Spanish and Taiwanese home cooks for what they share in their Whatsapp/Line groups. There are tons of written recipes and cooking videos that easily gets mixed among other chats.

If only there's a way to organise these gems better!

Project showcase laptop mockup
Project showcase laptop mockup
Project showcase laptop mockup

Recipe book 2 generations old

Project showcase laptop mockup
Project showcase laptop mockup
Project showcase laptop mockup

Cooking recipe in video format.

How can we make group interaction happen between users?

How can we make group interaction happen between users?

How can we make group interaction happen between users?

I ran a quick ideation session with a core team of 1 PM, 1 Analyst, 1 lead engineer and ran Impact x Effort analysis on each of the ideas to measure the result. The Cookbook idea first came up as a channel where family members can share cooking recipes and chat among each other.

We gave it broader scope to not just include families, but groups in general.

I ran a quick ideation session with a core team of 1 PM, 1 Analyst, 1 lead engineer and ran Impact x Effort analysis on each of the ideas to measure the result. The Cookbook idea first came up as a channel where family members can share cooking recipes and chat among each other.

We gave it broader scope to not just include families, but groups in general.

I ran a quick ideation session with a core team of 1 PM, 1 Analyst, 1 lead engineer and ran Impact x Effort analysis on each of the ideas to measure the result. The Cookbook idea first came up as a channel where family members can share cooking recipes and chat among each other.

We gave it broader scope to not just include families, but groups in general.

Ideation with storyboarding involved.

Validating with an MVP

Validating with an MVP

Validating with an MVP

I wanted to validate the idea very quickly before we decide to pour more effort. To do this, we launched an MVP in Taiwan with ±1000 users. We monitored the initial release for 2 weeks to evaluate the interest.

The MVP only enabled them to add recipes to cookbooks created by our staff, nothing more.

Click here to visit the MVP!

I wanted to validate the idea very quickly before we decide to pour more effort. To do this, we launched an MVP in Taiwan with ±1000 users. We monitored the initial release for 2 weeks to evaluate the interest.

The MVP only enabled them to add recipes to cookbooks created by our staff, nothing more.

Click here to visit the MVP!

I wanted to validate the idea very quickly before we decide to pour more effort. To do this, we launched an MVP in Taiwan with ±1000 users. We monitored the initial release for 2 weeks to evaluate the interest.

The MVP only enabled them to add recipes to cookbooks created by our staff, nothing more.

Click here to visit the MVP!

The MVP feature in white boxes

MVP results and next steps

MVP results and next steps

MVP results and next steps

Key engagement data from the MVP that we monitored:

• Average of ±30 recipes added to cookbooks / day

• Average of ±80 recipe comments added / day

We validated that idea and it's time to double down. I organised more interviews in between iterations to help decide which feature we build next.

Iteration - interview - iteration - interview…

06

06

06

Iterate, iterate, iterate

In one of the interviews I found out that the recipes people add to their cookbooks are always 1 of 3 possible kinds of recipes.

  • Recipes they’ve written themselves

  • Recipes they’ve cooked / tried at least once

  • Recipes they’ve bookmarked for future use

With this insight, I added tabs to the recipe addition flow to automatically recommend recipes they’ve interacted with. This fix raised the average number of recipes added per Cookbook by 366%!

The updated way to add recipes.

07

07

07

Final word

The project contributed to an average ±8% DAU growth in Taiwan and Spain.

As a result, we released the feature to the other countries with strong communities (Indonesia, Greece, Spain, Argentina, Uruguay). The end result is a global DAU growth of ±4.3% a month after release, compared to the 3 months we spent in Taiwan and Spain

What I would do differently:

  • To upscale to more markets faster. We waited for 4 iterations before we upscaled, while MLP features were done at the 3rd iteration.

  • To also consider other user groups that has potential to bring more users to the platform, such as cooking influencers.

Key screens.

Key flows.