Online marketing tool

Graduate Internship

My graduate internship saw me work with a design agency on an online marketing tool. The user base of the tool were the design firm’s clients, and the environment was their CMS.

The users are familiar with Google Analytics, but they find it difficult to use.

The goal of the project was to mitigate the client’s difficulties with understanding data from CMS.

UX research, UX/UI Design, Concepting, Prototyping and Testing

Eight months, complete design process in five sprints

Users are owners and marketing teams of small- to mid-sized companies

Understanding the problem

In order to understand why users have difficulties with analytics software, I researched the user base. I approached the research in a twofold way: by looking at the specific users, as well as general themes commonly encountered in CMS and analytics environments.

To learn about the specific users I was designing for, two interviews were conducted with owners of mid-sized companies. The interviews were focused on their use of web analytics. Additionally, I also got insights about user pain points from my agency’s marketing team.

To understand the general themes of issues, I did desk and literature research.

Through the research I found several pain points and behaviour patterns.

Firstly, the lack of knowledge. The use of KPIs, terminology, and interpretation of data results proved to be problematic for the users.

Secondly, the lack of awareness. Users didn’t know how and why to use analytics.

A behaviour pattern emerged, in which the user interprets the data, tries to find insights, makes decisions and actions based on those insights. Those actions include changing the website or launching promotions. Once the actions are completed, the data is analysed again in order to evaluate the success rate of the actions taken. Then, the process begins all over again. This created a core loop.

Process loop pattern

Problem statement

How do you make sure that users know what they can get out of webanalytics in order for them to create meaningful follow-up actions to improve their website?

I applied several visualisation methods to the insights from my research. Those included personas, empathy maps, and a user profile. They were continuously updated through the design process and served as a basis for the five design criteria the final product had to adhere to.

Users need to find valuable insights in a relatively brief time
Users need to base their follow-up actions on the found insights
Users of different knowledge levels should be able to use the system
Users need to feel confident using the system
Design agency touchpoints need to be incorporated into the design

Persona made as a user profile. Was continuously updated.

I used multiple diverging and converging concepting methods, including mind maps, longlist/shortlist-method, post-it method, sketching, low-fidelity prototyping, COCD Box, and focus group and prototype testing.

Post-it diverging concepting method.

COCD-Box Converging Concepting method.

Post-it diverging concepting method.

Six draft concepts were created with different bases and assumptions. Those were a personal assistant mobile app, a heavy gamification web application, an infographic-based email service, and a dashboard based on the user’s website layout.

Initial concept

I created a main concept loosely based on earlier drafts. This concept consisted of a dashboard with data collected by Google Analytics, presented in widgets. Users could set KPI goals to make data measurable. Gamification was incorporated to motivate users to actively use the dashboard and reward them for doing so. Microlearning also played a huge part in this concept; users would slowly get accustomed to KPIs and the overall experience of using analytics. This is done by gradually introducing terminology and features to the users. They also would start learning how to use this dashboard during kick-off meetings with online marketeers of the design agency. This created a service design.

Tests and Problem-solving

A prototype was made and tested with two participants: a product user and the CEO of the design firm. During the tests, I found out that the goal-setting function did not meet the user’s expectations. In my prototype, the goal-setting was performed by the user only; when in reality, it is mostly the online marketeer that does so. The setting in which the goals are set are kick-off meetings with the user and the marketeers. Therefore, I changed the user flow to reflect this: the goals were still a prominent part of the design, however, the goal-setting was now part of the “pre-experience”. Other feedback was that goals should be more visible in the user interface.

During the next four sprints, I used other methods to reshape and reframe this concept, including scenarios, storyboards, sitemaps, style guides, expert reviews, design principles, and multiple diagrams. Three other iterations were made to finetune the concept according to the constantly updated design criteria and problem statement.

Iteration 2

Iteration 3

Iteration 4

Iteration 5

Final Concept

The final concept was a one-page analytics dashboard which consisted of a library of information on all the different KPIs. The user would click on a prominent button of a specific KPI, after which a modal dialog would come up. It would show information about that specific KPI and its purpose for a successful website.

Goals were set during kick-off meetings with online marketeers. Those goals could be updated and added by using an intuitive contact form within the dashboard. This would increase the contact between the user and the online marketeer to discuss potential insights and follow-up actions. Goals are vital for users to maintain intrinsic motivation.

The core loop discovered in earlier research was used to create the final prototype, which was successfully tested. Users found it easy to understand and could find insights and were encouraged to act upon them.

The end result was a final working prototype made with Sketch and Flinto and a design rationale essay on why this concept solved the problem statement.

Final concept made in Sketch and prototyped in Flinto.

Styleguide made for the final prototype.

Key takeaways

I got the chance to be responsible for the creation of a product that would benefit users and clients. I learned how to deal with tight deadlines, multiple high-fidelity prototypes, and ever-changing design criteria.

The most valuable thing I learned from this project is that I have the ability to redesign a concept and product when constraints are tightened or a core piece of the concept doesn’t hold up in tests.

Although this was an individual project, I worked together with other designers and developers whom I counted on to help and advise me throughout this journey.