ClearGlass

ClearGlass Analytics is an independent data analytics company that enables asset owners, investment advisors, and asset managers to evaluate the Value for Money (VfM) of their investments through transparent cost and performance data.

As part of the product team, I contributed to the development of a SaaS platform designed to serve all sides of the asset management market, delivering structured cost transparency and advanced analytics across the UK and European investment ecosystem. The platform integrates industry frameworks such as the Cost Transparency Initiative (CTI) and the Cost Transparency Standard (CTS), enabling consistent data collection, benchmarking, and more informed investment decision-making.

Designing for this product presented unique challenges due to the highly cyclical nature of the industry. The platform operates around critical annual reporting cycles that determine when new features, improvements, and data submissions can be tested and adopted by users. These overlapping cycles created a distinctive product development rhythm and required careful prioritisation to ensure the platform evolved without missing key market opportunities. As one of the first platforms of its kind, the product played a pioneering role in bringing transparency and benchmarking capabilities to the asset management industry.

The organisation operated with a strong product-led mindset, with well-informed stakeholders actively collaborating across teams. Continuous experimentation, research, and rapid shifts in focus were embedded in the company culture, creating a fast-paced environment where multiple teams ran parallel initiatives to validate ideas and improve the product.

Conducting UX research in this domain presented additional complexity. Many users interacted with the platform only periodically during reporting cycles rather than daily, which made recruitment and engagement more difficult. To overcome this, we conducted regular user interviews (approximately every two weeks) and supplemented qualitative insights with internal data analysis and feedback from customer-facing teams. Asset owners were generally more open to participation than asset managers, which influenced our research strategy.

We triangulated insights from multiple sources to inform design decisions, combining user feedback with behavioural data from analytics tools such as Google Analytics, Hotjar, and Mixpanel. These insights were shared across the organisation on an ongoing basis, ensuring that product development remained aligned with real user needs and market dynamics.

Asset Manager Data Submission Journey

Designing a scalable ingestion and validation workflow for complex financial data

Overview.

ClearGlass’ platform relies on collecting detailed cost and performance data from asset managers, which is then processed and delivered to asset owners for benchmarking and reporting. Asset managers were responsible for uploading multiple datasets in different formats, often across several files. Because this data feeds directly into the platform’s analytics and reporting capabilities, the quality and structure of submissions had a direct impact on downstream processes. The project focused on redesigning the asset manager submission journey to improve data quality, reduce operational friction, and accelerate the delivery of investment insights to clients.

My Role

Responsibilities included:

  • Leading product discovery workshops alongside the product manager assigned to the project

  • Mapping complex multi-step user journeys

  • Designing submission workflows and complex validation mechanisms

  • Creating wireframes, prototypes, and high-fidelity designs

  • Running iterative testing and analysing behaviour through analytics tools (HotJar & MixPanel)

  • Collaborating closely with product managers, engineers, and customer success teams

  • Supporting continuous iteration across multiple product cycles

The Problem

The original data submission process presented several challenges:

  • Asset managers could upload multiple files in various formats

  • Spreadsheet errors were common and difficult to detect early - there was no automated validation and patterns listed out

  • Invalid data often required manual intervention from internal teams

  • Fixing issues late in the process delayed report delivery to asset owners

At the same time, we had to carefully manage the trade-off between data integrity and user effort. Asset managers interacted with the platform infrequently, often under time constraints, which made them particularly sensitive to friction and complexity. The challenge was to introduce more structured and reliable data capture mechanisms while preserving a streamlined, low-effort submission flow that would not discourage engagement.

Key Constraints

Industry reporting cycles
The asset management industry operates within strict annual reporting windows, meaning product improvements could only be tested and iterated during specific periods each year.

Diverse data formats
Asset managers submitted data in multiple spreadsheet structures and file types, creating a wide range of potential errors and edge cases.

Limited user access for research
Asset managers were relatively difficult to engage directly, requiring the team to rely on proxy insights from customer success teams and behavioural analytics.

Highly variable usage patterns
Larger clients submitted significantly more files and interacted with the platform more frequently, while smaller firms engaged only occasionally. The design needed to support both high-volume and low-frequency users effectively.

Research & Discovery

Direct access to asset managers was limited, so we relied on multiple complementary data sources to understand user behaviour:

  • Insights from customer success and account management teams

  • Behavioural analytics using Mixpanel and Hotjar

  • Observation of real submission flows within the platform

  • Internal workshops with teams supporting and meeting asset managers

These insights helped us identify the most common friction points in the upload journey, particularly around file errors, validation delays, and unclear feedback during submission. Because usage was periodic and traffic volumes relatively low compared to consumer faced products, analytics-driven iteration became a crucial part of the design process.

The Design Challenge

The primary challenge was to design a flexible submission system capable of handling many scenarios, including:

  • multiple file uploads

  • different data structures

  • incomplete submissions (which were still used and passed on to the other end) 

  • spreadsheet validation errors

  • varying levels of user experience

The workflow needed to remain efficient and intuitive, while also ensuring that data entering the platform met the required quality standards.

Iterative Product Development

Because of the industry’s reporting cycles, the solution evolved through multiple product iterations.

MVP – Initial Data Ingestion

The first iteration focused on solving the most immediate need:
allowing asset managers to upload data files directly into the platform.

This initial ingestion tool provided a structured entry point for submissions but still relied heavily on internal teams to detect and correct errors later in the process.

MVP+ – UI-Based Data Validation

The next iteration introduced real-time validation at the interface level.

Instead of relying solely on internal teams to identify issues, asset managers could now:

  • receive feedback on submission errors

  • correct mistakes before completing the upload

  • resubmit improved datasets directly

This significantly reduced the need for manual intervention and improved the quality of incoming data. We focused solely on those points instead of trying to craft the ideal user experience. 

MVP++ – Structured Question Framework

Further research revealed that many issues stemmed from unclear or inconsistent responses from asset managers.

To address this, we introduced a structured set of submission questions, designed in collaboration with internal stakeholders.

The goal was to:

  • standardise responses

  • reduce ambiguity in submitted data
    improve consistency across asset manager submissions

This iteration further improved the reliability of the data entering the platform. We identified the answer patterns and we were able to create the list of answers which significantly reduced the internal team efforts in understanding the external users issues. 

Post-MVP Redesign

After observing the workflow across several reporting cycles, we identified an opportunity to redesign the journey more comprehensively.

This phase involved:

  • mapping all possible submission scenarios and edge cases

  • documenting the complete end-to-end journey

  • designing workflows capable of supporting a broader range of file types and submission behaviours

Key design activities

Journey Mapping
 We mapped all micro-journeys within the submission flow to understand every possible outcome.

Concept Sketching
 Early sketches explored potential interaction models and submission structures.

Wireframing
 Multiple iterations of wireframes helped validate whether all scenarios and edge cases were properly addressed.

High-Fidelity Design
 The final designs were refined into detailed UI specifications to ensure smooth handoff to engineering and minimise ambiguity during development.

Impact

The redesigned submission journey produced measurable improvements across the platform:

  • Increased number of successful submissions 

  • Faster data upload and validation workflows

  • Reduced reliance on customer and data teams to resolve submission issues

  • Highly accelerated delivery of reports to asset owners

By improving the efficiency and reliability of data ingestion, the project contributed directly to faster reporting cycles and increased operational scalability for the platform.