What is the Hercules?
Hercules is a powerful data management tool that automates the often laborious and error-prone tasks of data integration, cleaning, and transformation. By doing so, Hercules becomes a time-saving solution for organizations, improving overall efficiency. It shifts their focus from mundane and time-consuming data processing to more valuable endeavors such as in-depth analysis and informed decision-making. Ultimately, this leads to more agile and data-driven insights.
Context
One of the most well-known (and, in my opinion, very controversial) statements in UX is "You Are Not the User." But what do you do when, in fact, you are the user? Hercules is a solution I created to help me tackle one of the most laborious and tedious tasks one can encounter in the tech world – Data Normalization. In this case study, I will explain the Lean Product Process I undertook to bring to life a solution that addresses the pain points I regularly faced when building analyses based on usability studies conducted on third-party platforms.
Feature Prioritization Puzzle
Before Hercules, the company manually processed data from multiple sources, including third-party software like Userzoom, MUIQ and others. This process was time-consuming, error-prone, and expensive, leading to inconsistencies in data analysis and decision-making.
An underserved market
The standardization of data for use in the Data Visualization portal presented a bottleneck for the company. It heavily relied on manual processes and the expertise of a single person for data normalization. This meant that when multiple projects needed to be included in the data visualization portal simultaneously, a bottleneck would arise, delaying one or multiple projects and impacting the workflow of other areas of the business. More critically, it affected the customer experience, causing delays in the delivery they were waiting for.
The Quest for True Value
Given that, operationally speaking, we couldn't resolve this issue with the third-party platform, nor could we increase the number of personnel working on this process, as it would enhance efficiency but also lead to inconsistencies in how data points were categorized. Ultimately, and most importantly, we needed a way to integrate all datasets into a single database to facilitate the company's Digital Transformation strategy.
Planning & Execution
Prioritizing the Features Set was crucial to meeting the underserved needs of the target audience. This was achieved by focusing on features that addesse their pain points. The prioritization process enabled me to establish a robust product-market fit, taking into account various factors, including business goals, market trends, technical feasibility, user impact, revenue potential, user adoption, customer retention, and resources and constraints.
To build and optimize the product, we chose to utilize the Product-Market Fit approach within the Lean Product Process framework. Embracing Agile methodologies, particularly Scrum, we dynamically adapted our approach, fostering adaptability and responsiveness to evolving project needs.
​
Methodology & Framework
Product Roadmap
To kickstart the Product Roadmap, we first performed a hierarchical decomposition of the project using the Lean Product Process framework. This method allowed us to break down the project into manageable components, defining its scope through task, subtask, and deliverable delineation. This approach facilitated effective planning, resource allocation, and progress tracking throughout the project lifecycle, aligning stakeholders, prioritizing tasks, and clearly communicating goals and milestones.
With team members spread across four different time zones, establishing a clear communication plan became paramount to our success. We focused on utilizing tools like Jira and Teamwork to track progress efficiently and keep everyone aligned. Weekly catch-ups emerged as essential meetings where we refined and validated requirements, ensuring everyone was on the same page.
Coordinating cross-functional teams
Key findings and insights
The research process enabled us to gain profound insights into the requirements and expectations of our target audience. By immersing ourselves in the intricate details of their needs, I identified the pain points and recognized the frequently neglected unserved needs of our target audience. Below, I have compiled a list of the most prominent pain points and underserved needs that were uncovered during the Empathize and early stages of the Define phase. The pain points and underserved needs were subsequently organized using the GUT priority matrix, prioritizing problems and demands based on their importance.
Users lack an understanding of the relative importance of features for different segments
Users struggle to locate industry-specific feature examples.
​
Users remain uninformed about the latest developments in the industry.
​
Users lack access to competitor customer areas (behind the login).
​
Limited information available about the impact of certain features on client acquisition and retention.
Understand the impact of specific features on client acquisition and retention.
Make informed decisions regarding feature prioritization and development by understanding what customers truly value in the product.
Tracking competitors' features, and monitoring market trends in real-time.
​
​
A way to access competitor customer areas (behind the login) as most tools available a primary focused on public sites.
Design question
How can we efficiently standardizes data from multiple sources, minimizes the reliance on individual expertise, and enables seamless integration into a single database?
Personas & Targeted customers
After conducting market research, we have identified three personas associated with our Problem space.
 |
|---|
 |
 |
}
For a deeper dive into the Personas, please click here.
Key Features and Innovations
Prioritizing the Features Set was crucial to meeting the underserved needs of the target audience. This was achieved by focusing on features that addesse their pain points. The prioritization process enabled me to establish a robust product-market fit, taking into account various factors, including business goals, market trends, technical feasibility, user impact, revenue potential, user adoption, customer retention, and resources and constraints.
Automatic Keyword-Based Categorization
Algorithm that automatically categorizes datapoints based on relevant keywords. It also allow users to review and adjust categorizations through an intuitive interface
Recurring Datapoints Identification
Pattern recognition system to identify recurring datapoints within datasets which provides configurable parameters for defining recurring patterns and alerts for new occurrences of recurring datapoints.
Automatic unwanted Datapoint Removal
Feature to easily flag and remove unwanted datapoints.
Provide manual and automated options for flagging based on user-defined criteria.
Include a version control system or an undo feature to maintain data integrity.
Data Integration
Create a flexible data integration framework that supports the seamless merging of different datasets.
Iteration & Prototyping
The iterative process played a crucial role in refining our product. We initiated with a Low Fidelity prototype on Figma, offering a basic framework to gather initial feedback and insights. Following the first round of interactions and validations, we progressed to develop a High Fidelity prototype.
​
Low Fidelity
High Fidelity
_edi.jpg)
Design-to-code efficiency
By choosing Figma, we enhanced product documentation for CSS and HTML specifications. Figma streamlined the process, enabling our Development team to generate CSS code snippets and export design assets with ease.
Why Figma?
Let's get interactive! Give the prototype a try on Figma.
Lessons Learned & Key takeaways
This project posed some challenges that prompted us to reassess testing methodologies and launch strategies, ultimately yielding valuable insights and key takeaways for future endeavors. Among the lessons learned, I believe three were particularly prominent:
AI Integration
Problem: The initial release of Helix utilized machine learning techniques to categorize instances and data points. However, this required significant effort to train the system for accurate identification and classification.
Solution: By integrating AI models, we can streamline the process, making it faster and more accurate. The platform review to incorporate AI capabilities is already underway.
Mapping unhappy paths
Problem: Although different scenarios were considered during the design process, it was discovered post-implementation that some use cases were missed.
Solution: Incorporate more real-world data during the testing phase to identify potential issues that may not surface when using only dummy data.
Data Security and Integrity
Problem: The data processed by Hercules was stored on the same server as other products like Helix and the Feature Library. This caused concerns about data protection and integrity since the Business Intelligence platforms that rely on Hercules’ data require live data fetching for visualization.
Solution: Segregate the data processed by Helix into a dedicated server, accessible only via remote networks, ensuring enhanced data security and integrity.
Products
Take a look at some of the products I've developed to get a better understanding of my work approach and the frameworks I apply.
