Onboarding

CASE STUDY

1 Product Manager

1 Creative Director

1 Designer

2 Engineers

1 UX Researcher

1 Animator

TEAM

GOALS

Promote push-notifications to increase open conversions by 20%
Increase our MAUs by 100%
Promote new location services, with 15% of fresh install opting-in.
Launch two weeks prior to the Nov 2020 General Election (2.5 months)

First Impressions

How might we increase our mobile monthly active users.

A large project was delayed due to the challenges of 2020, which gave the mobile team an opportunity to refocus our energies on the NBC News mobile app and how we could provide a better experience and increase our monthly active users (MAUs). Our mobile app visitors greatly outperformed web visitors in terms of return visits, page depth, and video consumption. Once new feature would be visitor onboarding.

Onboarding Tasks

Our stakeholder team, comprised of myself, a developer, our PM, research, editorial and insights began with sharing inspiration from other apps and web sites. We took a few minutes to sketch our ideas, and then voted for the features most likely to address our consumers requests.

I created many flow charts with the dev team to find a solutions possible within our time and resourcing constraints. Left: Creating a slimmed down flow to achieve our MVP. Right: This was a spike on a possible login. We will add accounts to future iterations to take advantage of larger, cross-brand NBCU experience.

Three areas were selected for the new onboarding experience:

1) Through usage and surveys, our visitors showed high interest in local news and weather. How might we deliver geo-targeted content?

2) A portion of visitors will never opt-in to notifications. How might we connect with them less intrusively while delivering content important to them?

3) Re-engage current visitors. How might we reconnect with visitors? The news cycle ebbs and flows, and usage habits change. A newsletter options is one alternative to push notifications. It provides room to go deeper on featured storylines or long-form content. It also provides us a method of reaching our consumers who are not as actively engaged. And once we implement topic favorites, we are then able to deliver content directly from the visitor’s customization.

Location Services

Giving users the option to share their location was the first step into delivering local content, a high priority for users of the News app. Our first step is to request localization and be able to send geo-targeted notifications. With IDFA also coming up around the same time we had to be cognizant of soon-to-come privacy regulations.

Lesson learned early on: with complex experiences that have endless variations, get your QA team involved early and as often as possible. By laying out flow we were able to work with QA to determine exceptions, and find edge cases sooner in the process. Pairing with developers also kept everyone on the same page. A developer that came into the project for the last month was able to review the requirements and understand the objectives.

Notifications

Push notifications are extremely effective in consumer rentention, yet a large majority of our iOS consumers declined push notifications. The default system dialog boxes requesting opt-in are distracting, inconvenient and often poorly-timed based on current actions of the visitor. It’s easy for muscle memory to default to “Deny”. Switching to a controlled ask I knew we could move the task into a space where the visitor is already in a decision-marking, not consumption mindset

Sample of an early prototype used in testing to evaluate vertical vs. horizontal scrolling and affects of animated background graphics

I know. Newsletters. Once the bane of my existence, I was now making the best case for them. I was drawn to the opportunity to provide the user with a newsletter that is customized to their preferences via ML or other passive personalization methods. This first iteration was step in that direction. I expected very few people to signup for yet another newsletter, but over 20% of new users opt-ing to receive the Morning Rundown.

Email communication and generating ideas on how they can be customized in context of the vendor’s services.

I switched responsibilities from designer to PM when the project's PM took a 3 month leave. There was little information as to how the actual data would flow, and a general assumption that the newsletter sign up could use integrations similar to Web, but that was not the case. I recruited our Business Intelligence team to jump in with their expertise on how to securely gather, pass along, remove user data and while respecting privacy regulations.

Post-production assets

The last facet that ran through the entire process was connecting with our broadcast post-production team for video and animated assets. They were able to speed up the process selecting the already-corporate-approved video assets for intro and outro. They also transformed a set of my animated icons into a more polished and delightful set for onboarding. I wanted a small touch of play to prevent the tasks from feeling like chores.

Animated icons from post-production keep a functional interface feeling more personable.

The MVP is highly functional, but will have a second round to include features and branding to bring it to the level of a premium product.

Results

As a combination of efforts between the teams involved in the original goal, we got within 80% of our monthly active user goal between the projects and features we released:

The rate of onboarding completion peaked at 97% and has maintained 80% post-election.
We saw a modest gain in iOS notification opt-ins. We regrouped to assess potential updates ranging from more effectively messaging the value-add, describing the regularity of alerts, and even the order in which they appear in the onboarding process. Next steps will be implementing topic preference and notification
We beat our estimate of opt-ins for newsletters during the onboarding processes by 4%, opening a door of communication.
The best success was with location services. Not only did we beat the expectation by 16%, but its given us insight that will change the priority we place on location-based content in the near future.

Sketches

One of the first steps in the design sprints were bringing various disciplines together, do a series of quick sketches and then swarm on the ideas that best address user needs.