Many companies struggle to make their data work. To shed light on this, we spoke with Gabor Harsanyi, Head of the Data Office at Ericsson’s Marketing and Corporate Relations unit! With years of experience in data strategy and AI, Gabor shares his approach to aligning data with business goals and avoiding common missteps.
In this interview, Gabor offers advice on focusing on the right priorities, clarifying roles, and managing change effectively. He also highlights how AI is changing data strategies, while stressing that human judgment and creativity remain essential. As a speaker at the Data Innovation Summit 2025, Gabor will share his thoughts on building data strategies that are impactful and sustainable.
Read on for Gabor’s insights on creating a data strategy that truly supports business needs, how AI fits into the bigger picture, and how to avoid common mistakes along the way.
Hyperight: Gabor, you’ve led initiatives in data strategy, including AI and data privacy ethics. Can you share your journey and what led you to your current role as Head of the Data Office at Ericsson’s M&CR unit?
Gabor Harsanyi: Absolutely. First, I think I need to elaborate on M&CR. This abbreviation stands for the Marketing and Corporate Relations function, which operates globally with 700 employees. It encompasses various functions, including marketing, communications, sustainability, organizational health and safety, and corporate relations. I joined the company four years ago, leading the function’s data and analytics team within a center of excellence.
After some reorganizations, I transitioned to the strategy team, shifting operational analytics to a sibling team within the Marketing organization and focusing exclusively on data at a strategic level. This was a game-changing shift in our approach, moving from a bottom-up to a top-down strategy development process. I believe this direction better aligns with the overall corporate strategy and is more in line with our objectives.
My approach to data products has also evolved: previously, as part of the CoE, I responded to stakeholder requests, whereas now I prioritize data product needs that contribute to business objectives and enable us to measure our OKRs.
Hyperight: Some organizations struggle with aligning their data strategies with broader business goals. What are some mistakes companies make when developing a data strategy? Why do they occur?
Gabor Harsanyi: The biggest mistake is what I just mentioned: failing to align the data strategy with the business strategy. All data products must deliver value and business impact – if they don’t contribute directly to business success, they should be deprioritized.
Another common misstep is starting the data strategy process in the wrong order. Many organizations begin with stakeholder interviews to identify use cases, define data products, and then build a strategy to support these use cases. The correct approach is the opposite: first, develop a data strategy that supports the business strategy, and then align with stakeholders on the data product needs that serve those strategies. When establishing a new data team, another pitfall is focusing on governance before strategy. If we begin at a granular level, we won’t be able to cascade down a coherent strategy. The first step should be appointing an experienced leader or employee who can drive the data strategy.
Additionally, strategies are usually updated on a fixed schedule (e.g., quarterly or annually). However, if our data strategy is a supporting strategy rather than a core one, its updates should be synchronized with business strategy updates rather than adhering to a rigid cadence.
Hyperight: In your experience, what distinguishes a successful, well-integrated data strategy from one that remains isolated and ineffective?
Gabor Harsanyi: While the fundamental principles of strategy apply to all domains, a well-integrated data strategy requires three key elements: prioritization, RACI (Responsible, Accountable, Consulted, Informed) clarity, and effective change management.
1. Prioritization ensures that we focus on use cases that align with business strategy.
2. Additionally, we must measure what truly matters, not just what’s easily available. If a crucial metric lacks measurement tools, we must invest in them. For example, if I want to lose weight, a measuring tape won’t suffice – I need a scale. Similarly, we must prioritize actionable metrics over vanity metrics that look good but lack business value. The RACI model is essential for clarifying roles and responsibilities. Strategy owners should be accountable for execution, while operational teams should be responsible for delivering results. A common oversight is neglecting the ‘Informed’ stakeholders – keeping them updated is crucial to maintaining executive buy-in. Even minor updates should be shared, making a well-designed stakeholder mapping indispensable.
3. Without proper change management, even the best strategy can fail. As an advocate of education, learning, and development, I believe upskilling and knowledge-sharing are critical components of change management. This is especially important if your data team operates in a support function rather than the company’s core business. If non-data employees remain untrained, all efforts risk being wasted. I hope this interview itself serves as a form of knowledge-sharing for many readers.
Hyperight: Your presentation at the Data Innovation Summit 2025 will draw an analogy between data-driven strategies and the need for a business-aligned approach. Can you elaborate on this concept?
Gabor Harsanyi: Leveraging the AI hype, I took inspiration from the Terminator movie series, where the humanoid machine describes itself as ‘living tissue over metal endoskeleton.’ My approach to data strategy follows a similar principle: establishing a solid, long-lasting framework – the metal endoskeleton – that requires minimal changes over time.
Meanwhile, the ‘living tissue’ represents the continuously evolving elements of the strategy: the data products. This aspect involves experimentation, iteration, and adaptation. If the business environment or strategy changes, our data use cases must evolve accordingly.
Hyperight: What steps can organizations take to ensure that their data strategy is aligned with real business needs rather than becoming an isolated technical initiative?
Gabor Harsanyi: The framework – the metal endoskeleton – should be robust and define the most critical, long-term aspects of the strategy. It serves as the foundation that evolves alongside the business strategy, always following one step behind. The business strategy comes first, followed by the data strategy, and then more granular levels, such as data products. A data strategy is not merely a technical initiative – it’s a communication tool that secures and maintains business buy-in. Technical aspects are just one component of data strategy, with the details covered in data governance documentation.
Two key elements keep a data strategy alive: change management and continuous communication. Regardless of execution progress, ongoing communication with business teams is crucial to maintain engagement and support. Without it, interest fades, and buy-in is lost.
Hyperight: Many companies struggle with stakeholder alignment, particularly when non-data teams feel disconnected from data initiatives. What approaches have you found effective in bridging this gap?
Gabor Harsanyi: This is where change management comes into play. The biggest challenge is that non-data teams often feel disengaged from data initiatives and may not fully understand data concepts. However, most people recognize that data is a valuable asset, fundamental for decision-making and AI applications. At the same time, there is fear that AI will replace jobs. The key message should be: ‘Let the robots take over the work, but not our jobs.’
To achieve this, upskilling is essential for both data and non-data teams. There are two types of education that must coexist: maturity-based training, which categorizes employees from beginners to professionals, and tool-based training, which focuses on the practical use of specific tools.
Hyperight: With AI playing an increasing role in data strategy, how should businesses approach AI-driven automation while preserving human insight and creativity?
Gabor Harsanyi: I anticipate a future where analytics will become a smaller component of D&A teams, while data roles will require more advanced skills. Generative AI can already produce insights, charts, and strategic analyses like SWOT reports. As a result, traditional data analyst and insight manager roles will become obsolete in their current form.
AI needs three things to deliver valuable insights: vast amounts of high-quality data and a ‘caretaker’ who provides and curates that data. These caretakers will be crucial for insight generation, requiring expertise in prompt engineering and subject matter knowledge to interpret AI outputs, correct errors, and mitigate hallucinations. This is where human insight and creativity remain indispensable.
Hyperight: What are some emerging trends you foresee in the intersection of AI and data strategy, particularly in ensuring data ethics and compliance?
Gabor Harsanyi: Some companies are creating dedicated AI spaces within their data strategies. My short-term approach aligns with this, but in the long run, I advocate for AI’s full integration into the data strategy framework. AI tools should be managed alongside other tools based on their purpose and processes rather than treated as a separate category.
Regarding ethics and compliance, generative AI introduces new considerations beyond traditional data ethics. My approach includes seven key areas:
- Privacy concerns (e.g., GDPR, CCPA)
- Business-sensitive data sharing
- Data breaches and misuse
- Content authenticity
- Intellectual property rights (IPR) management
- Harmful biases and discriminatory outcomes
- Sustainability
New challenges arise, such as AI generating PII from non-PII data or using copyrighted images, leading to IPR risks. Sustainability is another concern – AI, particularly in image and video generation, consumes significant energy, potentially leading to future “AI- shaming,” similar to flight-shaming if not properly managed.
Disclaimer: The views and opinions expressed herein are solely those of the individual and do not necessarily reflect the views or opinions of the company.
If you’re looking to dive into how AI is shaping data strategies and the role of human insight in an increasingly automated world, don’t miss Gabor’s session at the Data Innovation Summit 2025! He will explore how companies can create data strategies that truly align with business goals and adapt to the rapidly changing technological landscape. Gabor will also provide advice on navigating challenges like stakeholder alignment and integrating AI into the broader data strategy.
Join Gabor as he shares his approach to building data strategies that are technical, sustainable, and effective in driving business impact. Whether you’re a data professional or a business leader looking to better leverage data, his session will offer insights into how to keep your strategy both agile and aligned with your organization’s objectives.
Add comment