Manga Rock: Campaigns


For our internal content managers and executives who need to drive user engagement within the Manga Rock app, Campaigns is a new feature on Manga Rock Addons – the management system for all non-core content – that allows its users to create and run effective content targeting campaigns.

Unlike previous attempts, Campaigns offers a user-friendly interface, built-in analytics, and supports multiple channels.

I led a team of two designers on its conceptualization and design.

My Contribution

Concepts & Ideation
Information Architecture
Enterprise UI/UX Design


2 designers
1 product manager
4 engineers



Problem Statement

Previously, we have launched the For You feature, allowing the content curators to control the content shown to end-users in the For You tab inside the MR Comics app on iOS, Android, and Web.

However, some of the feedback we got is that the current “For You” tab in the app right now is not true to its meaning, as what the content team has to do is configuring the sections shown each and every single day, which is time-consuming and just plain absurd. The only way they can “truly” deliver content to a targeted set of users is through push notifications, albeit it is a very manual and tedious process in which a list of users must be generated by the BI team.

The need for targeting is very apparent. We recently opened the doors for premium content onto Manga Rock. To cover the procurement cost we spent on this content and maximizing the profits, it is important for businesses that we can successfully sell them to potential buying users.

There are many ways to sell things, however, we don’t want to settle with spam and bombardment, which would result in a negative experience for the end-users. After some discussion, we decided on a targeted approach, which, like targeted ads, will only be shown to you if we deemed that you would be interested in the content. This would serve both the business goal and the user’s needs, as they are exposed to content of value to them.

With this approach, we need a system where our content team can plan and execute their content targeting effort.


The users for this feature is content team members such as content managers, content executive, content curators, etc… tasked with conducting the content targeting effort.

From talking with our content team members, we discover that, in general, our content team has many items in their inventory that they want to drive sales, so they want to deliver them to users who would want to consume this content.

Sometimes, these efforts coincide with important occasions and festivities, as well as to meet with quarterly KPIs, etc… and because we can’t expect people to work 24/7, the ability to plan ahead and schedule is also important.

Naturally, after creating a lot of these targeting efforts, things would start to get out of hand, so they would want to manage them in an effective way. Also, the ability to measure the performance and compare the results of different efforts is also important in ensuring the effectiveness of the operations.


The goal is to enable the content team to effectively deliver the right content to the right users.

To accomplish this goal, we decided to:

  1. Give content team the ability to:
    • package content in a way that is appealing to users
    • select the channel they want to deliver
    • pick the target audience they want
    • schedule the targeting to specific dates
    • compare the performance of various targeting campaigns
  2. Remove the need to manage for you releases everyday

In doing so, we bring about the following benefits:

  • For our end users, they can now enjoy content tailored to their taste.
  • For external publishers, their content can get more reach and they can get more potential revenue.
  • For our content team members:
    • Able to curate content and deliver it effectively – this keeps user inside the app and fulfills Manga Rock’s original promise to its users
    • Able to sell content effectively – the more potential users are exposed to targeted content, the more likely they are going to pay for it
    • Better understanding of users & content relationship – our content team will gain more insights and get better at understanding user’s interests, these data will eventually be fed to our data layer
    • Focus on content strategy – now content team can focus on each content instead of the tedious configuration of for you structure

However, it also creates some downsides:

  • The content team loses the ability to control exactly how users see the for you → however, it is a good thing, since eventually we want our machine learning to take care of what to show to users.

Design Direction

We want content team members to be able to use this system in a way that promotes effectiveness. To accomplish that, the design has to be:


The first characteristic that defines effectiveness is accuracy. The targeting must work as the content team member intended it to be. The correct content must be delivered to the correct audience. The number of views and clicks must be accurately tracked and shown. These data will not only serve to inform content team members to come up with subsequent tries, they also are the foundation data that eventually going to be fed to our machine learning and serve as the material for our important features such as predictions, suggestions…

Most of the work that contributes towards being accurate relies on ensuring our backend systems work effectively and harmoniously with each other, without problems.

It should be noted that being accurate does not necessarily mean data has to be displayed in real-time, a slight delay is acceptable, the only thing that is required is that the number is accurate at the point the data is populated.


Information should be made visible and easily accessible to the system users. Because it is up to them to draw insights and conclusions from the data collected from these campaigns. In this phase of targeting, we rely heavily on human intuition to do all the comparisons and experiments with different content, audience, and channel combinations. Therefore, it’s important that we provide them the convenience and abundance of information for them to decide upon.

Some information a content team executive care about:

  • Number of views (reach)
  • Number of taps/clicks/action (engagement)
  • The conversion rate
  • The audience size
  • Number of grow over time
  • Historical data
  • etc…


The thing about doing campaigns is that much of it is about trials and errors. By experimenting with different content and audience and channels, our content team gets better and better at coming up with future campaigns. They will get better at understanding our user’s interests and the affinity of each content that is given to them. In order to achieve this aim, lots and lots of campaigns will be created and tested. For every content or audience, multiple campaigns will be running at the same time, because the only way to make sense of the data gathered from campaigns is by comparing them with each other.

So we can be sure that there will be a huge number of campaigns being created and running on the system. Therefore, the ability to manage these campaigns becomes very important. The content team needs to be able to organize, search, filter extensively in order to drill down to the info they want and extract meaningful insights from them.


Next, I came up with conceptual models to meet the content team’s needs.


This is what we decided to call a targeting effort that the content team created. Based on the need of the content team regarding the different methods of delivery, we divided the campaigns into 2 types:

  • Push delivery – channels that deliver content one-time directly to the user, through their device or inbox
  • Pull delivery – channels that dynamically resurfaces the content over a period of time at various touchpoints in the user’s discovery and browsing flow inside the app

The difference in campaign types lies in the way content team user can select the campaign date:

  • For Push Delivery campaign, content team selects the release date when the content get sent out to users.
  • For Pull Delivery, content team selects the start and end date when the content is visible to users.

A campaign is consisted of:

  1. Content oid – the oid of the series, collection, artwork, sticker packs…
  2. Audience oid – audience object from Audience MS
  3. Channel – for you, new, browse, reading break, push notification, email…
  4. Start & end date –  (for Pull Delivery) the running date for the campaign
  5. Release date – (for Push Delivery) the date the content get sent out
  6. Status – track the state of the campaign: draft, pending, ongoing, completed, archived
  7. Tags – to manage the campaigns
  8. Metrics
    1. Reach – number of times user get exposed to the content
    2. Engagement – number of user interactions upon the campaign
    3. Conversion Rate – the percentage of engagement / reach
  9. Extras – additional data such as push notification payload


Define the placement of the campaign in the apps. A campaign can target multiple channel. Each channel can support a certain number of content types.

The table below shows which type of content can be shown in which channel.

For You✔️✔️✔️
Reading Break✔️✔️
Push Notifications✔️✔️✔️


Used to attach to each campaign as a way to manage them more effectively. A tag contains a name and description.

User Journey

To illustrate the many systems involved in making this feature possible, I made a sequence diagram of the step-by-step process regarding how a content executive deal with campaigns.

Information Architecture

Based on the various touchpoints discovered in the above flow, I came up with the following information architecture.

Below is a sitemap of how these screens will fit into the bigger picture of the Addons product.

User Flows

Visual Design

This is how the design looks like after several rounds of design iteration and implementation.


After launching in 2018, content team members made use of the feature to run multiple campaigns. As a result, user engagement in the app saw a significant increase.