CASE STUDY: Creating a Data Driven Culture at TD Ameritrade with Derek Strauss

Executive Summary

In this webinar, Derek Strauss shares his five-year journey as the first Chief Data Officer at TD Ameritrade, detailing how he transformed the organisation into a data-driven enterprise. This session provides invaluable insights into building sustainable data capabilities, overcoming organisational resistance, and creating measurable business value through strategic data initiatives.

Webinar Details

Title: CASE STUDY: Creating a Data Driven Culture at TD Ameritrade with Derek Strauss
Date: 2025-08-07
Presenter: Derek Strauss
Meetup Group: DAMA SA User Group Meeting
Write-up Author: Howard Diesel

Introduction and CDOIQ Background

Derek Strauss opens the webinar, a distinguished South African data professional and the inaugural Chief Data Officer at TD Ameritrade. He shares that he will discuss the transformative role of the Conference on Data Quality (CDOIQ). Founded by Professor Richard Wang at MIT, the CDOIQ has evolved significantly over its 19 years, expanding from a modest gathering of five members to a vibrant symposium attracting over 2,200 attendees worldwide. This growth includes regional conferences that further emphasise the global commitment to data quality and governance.

During the 2024 symposium, there was a palpable energy surrounding artificial intelligence and large language models, with organisations increasingly aware of the necessity to enhance their data foundations before launching AI initiatives. Derek then stresses the importance of the CDOIQ in guiding organisations through the complexities of implementing AI while maintaining robust data governance and quality controls. This critical focus helps to mitigate the risks associated with poor data practices, ensuring that organisations are better equipped for the future of AI development.

The CDO Role – Tenure and Challenges

The brief tenure of Chief Data Officers (CDOs), which typically lasts only 18 months to two years, poses a significant challenge in the data industry, hindering their ability to implement meaningful, long-lasting change. Derek’s nearly five-year tenure at TD Ameritrade stands out as a rare exception, largely due to his collaboration with a Chief Operating Officer who recognised the strategic importance of data. This executive played a vital role in shielding Derek from premature political interference, allowing him to concentrate effectively on executing data initiatives.

Derek underscores the necessity for CDOs to have strong executive sponsors who possess a deep understanding of data’s potential and can provide essential support in navigating organisational challenges. The presence of a knowledgeable champion not only enhances the likelihood of successful data initiatives but also fosters the longevity of the CDO role. Ultimately, cultivating such executive support is crucial for enabling CDOs to drive substantial change within their organisations.

Figure 1 Case Study: Creating a Data Driven Culture at TD Ameritrade

Understanding Business Drivers and Strategy Alignment

The Gavroshe Seven Streams (G7S) framework, introduced by Derek, represents a strategic maturity assessment tailored for effective communication at the C-suite level. This framework is the culmination of decades of collaboration with industry pioneers such as Bill Inmon, Claudia Imhoff, and John Zachman. Derek argues that for data strategies to be truly successful, they must be deeply embedded in the overarching business strategy rather than merely responding to fleeting technology trends.

Derek warns against the common pitfall of organisations chasing “silver bullets” like metadata repositories, data warehouses, data lakes, and emerging AI solutions, which vendors often tout as comprehensive fixes for all data-related challenges. He highlights four critical success factors: alignment with business goals, identification of the primary business driver, sustainability of business value beyond initial implementation, and scalability of solutions. Without these foundational elements, data initiatives risk devolving into isolated projects that lack perceived business value and are ultimately at risk of discontinuation.

Figure 2 About Gavroshe

Figure 3 A Perspective from DW2.0

Building the Data Strategy – Short-term Wins and Long-term Vision

On his first day at TD Ameritrade, the CEO presented Derek with a bold challenge: to transform the company into a more “Amazon-like” organisation that prioritises personalised, customer-centric experiences. This mission required a fundamental shift from an account-centric model to one centred on customers, especially given the company’s history of growth through acquisitions, which resulted in fragmented customer data. To support this initiative, the CEO pledged unprecedented funding directly to Derek, highlighting the importance of data transformation at the board level.

Recognising the need for a strategic approach, Derek’s team crafted a rallying cry to make data “fast, easy, and intimate,” while also ensuring it remained “accurate, accessible, and actionable.” By focusing on customer master data management (MDM) as a foundational element, the strategy enabled capabilities such as household aggregation, propensity modelling, event recognition, and personalised offerings. This clear vision not only addressed immediate business needs but also established a sustainable framework for future growth, ultimately positioning TD Ameritrade for scalable success.

Figure 4 “It’s Raining Silver Bullets. Let’s get it Right this Time.”

Figure 5 Outline

Data Architecture Vision – Customer-Centric Enterprise

Derek’s architectural vision aimed to revolutionise customer master data management by supporting intricate household relationships, enabling clients to belong to multiple households while retaining joint account ownership. Recognising the significant limitations within the organisation’s current fragmented systems, Derek identified the pressing need for cohesive customer integration, which was lacking in their operations.

The existing setup did not allow for linking multiple accounts to a single client or managing cross-household relationships, which hindered the overall client experience. His solution focused on targeting specific, measurable goals, including generating significant net-new assets, refining offer targeting and timing, and delivering personalised services.

To achieve these objectives, Derek pinpointed a data team that was overwhelmed and demoralised due to poor data architecture and excessive demand for their services. He discovered that the only viable asset was the Netezza-based data warehouse, as everything else was in disarray.

By prioritising customer-centricity and complementing ongoing initiatives, Derek’s team developed a robust master data management solution. This new framework not only addressed immediate data challenges but also laid the groundwork for enhanced business capabilities across the entire enterprise, ultimately improving the client experience and operational efficiency.

Figure 6 Developing the Data & Analytics Strategy – a Long-term Program with Short-term Wins

Figure 7 “Become more Amazon-like”

Figure 8 Intimacy without Creepiness

Figure 9 The Business Vision

Figure 10 The Role of the Chief Data Officer

Organisational Structure – The Four Pillars

Derek effectively established the Enterprise Data and Analytics Group (EDAG) as a critical business function, positioning it alongside the CIO and CTO under the COO. This strategic alignment was essential for recognising data as a valuable business asset. The organisation was structured into four pillars—two focused on business (Data Governance & Demand Management, and Analytics) and two on technology (Data Architecture & Administration, and Data Resource Development Centre)—to create a cohesive data management framework.

Within this framework, the governance and analytics pillars served as centres of excellence, each staffed by approximately 10 experts who coached data stewards across business units. This model facilitated solid-line reporting to business units while maintaining a dotted-line connection to the Chief Data Officer (CDO).

Derek then highlighted that CDOs with direct authority over architecture and governance achieve faster results and drive cultural transformation more effectively than those dependent on external teams. Ultimately, this structure successfully balances centralised control over essential functions with active business engagement, fostering a data-driven culture across the organisation.

Figure 11 Chief Data Officer – Position in the Executive Suite

Figure 12 CDO Organisational Components

Figure 13 CDO Organisation – Data Governance and Demand Management

Figure 14 CDO Organisation – Competitive Analytics Centre

Figure 15 CDO Organisation – Data Architecture and Data Administration

The Seven Streams Framework

The Gavroche Seven Streams framework is a vital tool for Chief Data Officers (CDOs) that identifies seven essential levers for effective data management. These include Data Governance, Data Architecture, Data Asset Development, Data Quality, Data Context (metadata), Analytics/Data Science/AI, and Infrastructure/Platforms.

Derek illustrates this concept with the metaphor of a starship captain monitoring seven dials, emphasising that success in data management hinges on balanced investment across all streams. A singular focus on governance and architecture, while sidelining quality, metadata, analytics, and infrastructure, can lead to failure.

When Derek began his tenure at TD Ameritrade, the organisation had a maturity level of essentially zero across all streams, except Data Asset Development, which was at level one due to the existing Data Warehouse. The desired target state involved progressing primarily to level three, with certain areas reaching level two. This gap analysis directly informed the data strategy’s seven components, aligning each with a specific stream and establishing a comprehensive roadmap for capability development.

Figure 16 Highest Paid Person’s Opinion

Figure 17 Data Management Mission Map

Figure 18 Data and Analytics

Technology Architecture Evolution – Beyond Silver Bullets

Derek presented a compelling critique of the traditional “data lake” model, proposing the innovative concept of a “data marshalling yard” to improve communication with C-suite executives and foster better organisational understanding. He argued that the data marshalling yard serves as an effective staging area where data from various sources can be profiled, analysed, and prepared for subsequent use, offering a practical alternative to the metaphor of abstract “lakes.” This approach also acknowledges the unique needs of different user archetypes, such as farmers, operators, and tourists, who depend on ACID-compliant databases that traditional data lakes cannot adequately provide.

By preserving essential components like the enterprise data warehouse and operational data stores, Derek’s architecture seamlessly integrates unstructured and semi-structured data into a cohesive framework. A key innovation in this model is the use of “textual ETL,” which enables the integration of semi-structured data into structured data warehousing for comprehensive analysis. Grounded in Bill Inmon’s DW2.0 concepts, this strategy effectively balances the adoption of new technologies with the retention of proven capabilities, enabling organisations to navigate the evolving data landscape without sacrificing functionality.

Figure 19 Data and Analytics – Key Initiatives

Figure 20 Business Competencies

Figure 21 Business Capabilities

Figure 22 Conceptual Data Subject Areas

Figure 23 Business Competencies – Data Entity Mapping

Figure 24 Supporting the Business Capability Map

Figure 25 Ensuring Sustainable Business Value

Figure 26 The Strategy should Address Several Aspects, including:

Figure 27 Data Ownership – Assignment Process

Figure 28 Data Ownership – Assignment Outcome

Data Governance Implementation

Data governance at TD Ameritrade uncovered significant quality issues, prompting the organisation to review its data management approach. An initial analysis of the “state” field revealed a startling 128 values, rather than the expected 50 US states, including erroneous entries such as “Singapore.”
This revelation not only shattered any denial regarding data quality problems but also galvanised support for governance initiatives among senior leadership. Derek’s team effectively mapped data subject areas to business ownership, illuminating the complexities inherent in customer data governance.

The intricacies of data ownership were particularly evident in the “client” subject area, which required a governance matrix to delineate responsibilities. This matrix identified a primary owner from the back- or middle-office, along with secondary owners from various divisions, including marketing, retail business units, and compliance.

Although this structure was complex, it accurately reflected the multifaceted nature of a multi-line organisation, in which no single unit has complete oversight of customer data. By combining subject-area governors with element-level owners, TD Ameritrade established a framework that fostered accountability and ensured a comprehensive data governance approach.

Figure 29 Subject Area Model

Figure 30 The Gavroshe 7 Steams (G7S) Playbook™

Figure 31 Gavroshe 7S Framework for Data and Analytics™

Figure 32 ‘The Wheel’

Figure 33 Gavroshe 7S Framework for Data and Analytics™

Figure 34 7SMM™ – Target State

Figure 35 7SMM™ – Target State pt.2

Figure 36 The Gavroshe 7S Play Book™

Figure 37 Typical Play Card in the 7S Play Book™

Figure 38 Designing a Robust Long-term Architecture

Figure 39 Key Architectural Questions

Managing Organisational Change

Derek encountered significant resistance from both the business and technology sectors while advocating for data governance within the organisation. Business leaders were sceptical about the need for data governance, viewing it solely as a technological responsibility, and they resisted the idea of their analytics teams reporting to a central centre of excellence.

At the same time, technology teams, entrenched in traditional waterfall methodologies, were wary of adopting cloud computing due to perceived risks in the financial services sector. The security and accessibility of data created further challenges, as data had been locked down so effectively that analytics teams struggled to access the information they needed.

To navigate these obstacles, Derek turned to two key resources: John Kotter’s “Leading Change,” which provides an eight-step process for effective change management, and Patrick Lencioni’s “The Advantage,” which outlines a framework for organisational clarity. Kotter’s model offered a structured approach to implementing change, while Lencioni’s insights helped the organisation define its identity, objectives, and success strategies.

This clarity facilitated effective communication through roadshows, allowing the organisation to better understand EDAG’s value proposition and rules of engagement. Ultimately, the culture change that Derek championed was intentional and systematic, laying the groundwork for lasting transformation within the organisation.

Figure 40 Why and Who

Figure 41 ‘Tourist’

Figure 42 ‘Explorers’

Figure 43 ‘Miners’

Figure 44 Evolving the Analytics Communities and their Tools

Figure 45 The DW and the Physical ODS

Figure 46 Data Quality Improvement (for Critical Data Elements)

Figure 47 Getting Value from Unstructured Data

Measuring Success – Beyond ROI

Derek critiques traditional ROI metrics as inadequate for measuring success, advocating instead for capability-based metrics that prioritise tangible business outcomes. His approach centres on the question, “What can I do today that I couldn’t do yesterday?”

This shift emphasises measurable advancements in business capabilities, including new functionalities enabled, delivered business value, improved agility, and the elimination of technology-related pain points. TD Ameritrade exemplifies this methodology by monitoring short-term wins, such as code master data implementation and data virtualisation deployments, as well as the initial components for textual analytics across various communication channels.

The longer-term initiatives at TD Ameritrade build on these foundational successes, reinforcing the importance of qualitative improvements that business users can clearly articulate. While quantitative scores may still be tracked, the primary focus remains on business-friendly capability enhancements.
This strategic emphasis has not gone unnoticed; it culminated in the prestigious 2020 International Society of Chief Data Officers Award. The recognition serves as a validation of the team’s efforts, highlighting the effectiveness of prioritising business outcomes over conventional technical metrics.

Figure 48 Integrating Structure and Unstructured

Figure 49 Other Components in the DW2.0 Ecosystem

Figure 50 Data Warehousing 2.0

Figure 51 Big Data – Technology Drivers

Figure 52 The Power of Data Science is Unleashed

Figure 53 Beyond DW2.0 – Has Big Data made DW and DQI Obsolete?

Figure 54 Laying the Capability Foundation

Figure 55 Organisational Change: “Change is Good!”

Figure 56 ‘Leading Change’ by John P. Kotter

Figure 57 ‘The Advantage’ by Patrick Lencioni

Figure 58 Measuring Success

Q&A Session – Practical Insights

In the Q&A session, key insights were shared on implementing effective data governance and architecture. Derek highlighted that Chief Data Officers (CDOs) who have direct control over architecture and governance tend to achieve substantially better outcomes than those who rely on external teams.
The discussion also emphasised the importance of upskilling existing talent and properly defining roles within the data governance framework. TD Ameritrade’s innovative approach leveraged frustrated talent with existing skills and applied a “Shark Tank” model, in which university research labs analysed anonymised data in AWS and proposed insights.

This method ultimately led to the selection of three universities for continued collaboration. Additionally, the concept of data stewardship was addressed, indicating that these roles were formalised through appropriate job descriptions and training.

Rather than creating new positions, this approach recognised and empowered informal work already being done in data profiling and metadata management. Overall, the session underscored the significance of strategic talent management and clear role definitions in fostering a successful data governance environment.