Brandtrack

🧢 Role: Product Designer, Researcher, Data Visualization Expert

🧠 Demonstrated Skills: Analytics, User Research, Prototyping, Testing, Design

⏳ Duration: 6 months

Overview

Pixability was the first company to be certified as a YTMP member in 2016. Since then, the collection of public YouTube data has helped inform best practices in campaign strategy. As a part of the company's transition to creating a self-serve platform, BrandTrack was conceptualized to offer a one-stop packaged solution for clients to access this data in a meaningful way. The goal of this initiative was to design a dashboard experience that captures YouTube insights where advertisers and campaign managers can get a holistic understanding of their respective YouTube ecosystem.

My Contributions

Sitting within the analytics team, I served as a hybrid between the analysts and the product designers. With the first version of BrandTrack released by the analytics team, it started off as a hefty data dump with repetitive tables and a small regard for a non-analytics user experience. From re-segmenting the data, to conducting user interviews and rapid prototyping, I co-led a team of 5 through the project from re-conceptualization to productization over the course of 6 months.

Business Outcomes

• Secured $85K in revenue across 20 clients upon initial launch as well as 2 partnerships

• Reduced analyst workload by 30%

• Established data visualization guidelines for organization to be leveraged for reporting, marketing materials, and more

Defining the problem

Our users are not actually analysts

When I was hired in February 2019 as the Lead Data Insights Artist (fancy title for the dedicated product designer for the analytics team), internal analysts had already come up with some version of what this dashboard could look like. The issue was… it was just that - a dashboard of numbers. I was tasked with transforming this into a real user experience - a dashboard that clients could use to inform their own next steps.

So why are we throwing all of these graphs at them?

The initial versions of BrandTrack included repetitive tables and visualizations, random use of brand colors, and an overall lack of fluidity from page-to-page. When scanning through the product, users would often second-guess themselves, thinking that they had already seen a visualization or set of data as there seemed to be no differentiation between elements.

PROS

• All of the data was put together
• Great for analysts

CONS

• Repetitive visualizations
• Data dump
‍• Lack of a story
‍• Filters and interactions are unclear
• Bad for marketing to executives

Segmenting the pieces

Finding the right story

The lack of a story was the biggest hurdle to overcome. There hadn't been a clear discussion about the goals of the product. The analytics team just knew they had all this data and they wanted to put it together. It was at least organized in the sense that similar data was with similar data.

However this creates a problem.

I like to call this the surplus of supplies dilemma.

The solution....If we equate the different segments of data to different types of art supplies, we can combine different pieces/tools to create an artists' paradise - the goal is to equip the user with various tools so that they can accomplish what they need! Let's create the perfect toolset.

Methodology

Defining Goals & User Testing

I started out by summarizing what each chart was doing - what information was it displaying and what purpose does it serve or could it serve for the user. I shuffled some of the pieces around and also went back to our executive team to write down major questions/problems they are hoping to solve with this product.

Once I gathered all of my information, I shuffled the pieces together and sorted them into stories. From there I was able to put together my wireframes....

Low-Fi Prototyping

Given the constraint of the product's "soft release" being built in a third-party data visualization tool, there were some limitations in design and how it translated into the program.

Hi-Fi Prototyping

After several iterations on the layout, in addition to the rest of the pages, we almost would not be able to recognize the original mockups. As a team, we came a long way in understanding thoughtful brainstorming steps and ideation.

Outcomes

Data Visualization Guidelines

While BrandTrack is its own product sitting within the Pixability ecosystem, it provided an opportunity to re-evaluate branding in the context of data visualization and insights. I developed a short data visualization-specific brand guide to assist my team in conceptualizing future work.

Learnings

Fight for good UX

As much as clients and stakeholders will keep adding to their wishlist, often times they don’t know that they are asking for clunk. My fight for good UX was focused on streamlining their thoughts into simple stories that inspire rather than cloud their minds.

There is a learning curve for both parties

We had to employ tangible activities to really help stakeholders sort through their requests. This helped them see the problem that we were addressing and smoothed out our process moving forward.

Adapting to changing requirements

Documenting every meeting was crucial. Reminding the team of what was agreed on in previous meetings helped to eliminate repetitive work and drill in best practices around our product development.

A New Take

After BrandTrack was released into market in early 2020, a few months had gone by and I wanted to revisit the project. When working on it in its first release, the main project restraint was that it had to be built within another system (there was no engineering support to build out the front-end experience!) I wanted to reimagine BrandTrack in a world where it could be built from scratch.

Personalization

After release, there were user retention issues - one of the many points of feedback our team received was the request to save progress from the user's previous session.In this redesign, I brought in more human-centered elements, welcoming the user on the landing page, and reserving that top-left corner for specific-to-the-user metrics and information.

More data at a glance

In the released version of BrandTrack, the user has to heavily rely on filtering to get to different metrics for the same data points.Often when I am doing my own analyses, I will also look at four different scatterplots to find trends in data.