Prashanth Rajendran, Product Manager, Samsung Research America
During a recent earnings conference call, Microsoft CEO Satya Nadella observed: “Every AI application starts with data and having a comprehensive data and analytics platform is more important than ever.” Although much has changed over the past year with the emergence of generative AI, data quality remains a fundamental part of any business – essential not only for modern product development and data management. customer experience, but also to LLM use cases such as augmented generation by recovery (LLM RAG).
After a fun afternoon of planning that conferences on generative AI in 2024 To attend, I recently caught up with Prashanth Rajendran – head of data products at Samsung Research – about his experience building data strategies for various organizations, from industry giants like Samsung to fast-growing startups like Twilio and Cirrus Insight. We’ve explored some high-level ways businesses can think about their data strategy and leverage data in ways that drive growth and success.
How do you think about data strategy at a high level?
At its core, a data strategy is a fundamental plan that defines how an organization will manage its data assets to achieve its business objectives. This encompasses the people, processes and technology needed to effectively capture, store, process and use data.
A good data strategy should be aligned with the overall business strategy and take into account the organization’s current data landscape, future data needs, and potential risks and challenges. When developing a data strategy, I have four key considerations businesses should keep in mind.
The first is define the data vision. What do you want to achieve with the data? This could include improving decision-making, creating new data or AI-driven products, or improving customer experience.
This then involves assessing current data assets and identifying gaps.. Do you have the right data to realize your vision? This includes considering the types of data you have. Is there any key data you don’t have? This may include data on customer behavior, market trends or operational performance.
Then you must determine the appropriate technology and tools. Do you have the right building blocks to collect, store and analyze your data?
Finally, you must build a data-driven culture. Do you have the right people and processes in place to leverage data effectively? This could include creating a culture of data-driven decision-making.
Overall, a well-developed data strategy can help organizations make better decisions, improve operational efficiency, and drive growth and innovation. By taking a holistic approach to data management, businesses can unlock the full potential of their data assets and achieve their business goals.
How can companies create a data moat?
There are several things to consider when it comes to getting the “right data” based on your context.
The first is to identify the right data source. Not all data sources are equal. Public data sources such as the U.S. Census or economic data may be readily available, but they may not provide a competitive advantage to your application. Anyone can create the same AI application.
On the other hand, first-party data – proprietary or proprietary data, like the number of people viewing a real estate listing on online aggregator sites like Zillow or Redfin – can provide a strong competitive advantage for your product .
To create a data gap, you can focus on first-party data, such as: mobile/web app usage data, e-commerce transaction data, and behavioral data such as store visits and customer reviews.
Another way to create a data gap is through enrichment, which involves supplementing missing or incomplete data with external sources. For example, if you have data on existing customers, you can leverage third-party data sources to enrich the data with “employee count,” allowing you to track customer headcount growth and potentially initiate an upsell conversation.
Another key consideration is prioritizing data sources. This involves considering factors such as the cost of acquiring and processing data, the potential impact on your business, and the effort required to clean and integrate the data.
However, it is important to note that identifying and prioritizing the most valuable data sources is only the first step. Once you have identified these sources, a significant data engineering effort is required to clean the data as needed. This is why successful data managers often say that 80% of their efforts are spent identifying, modeling, extracting and processing data, rather than building ML models.
By focusing on the right data sources and prioritizing your efforts, you can create a competitive advantage that sets your product apart from others in the market.
How do you see companies effectively measuring the return on investment (ROI) of their data strategy?
There is no one-size-fits-all approach to data initiatives, and businesses must tailor their strategies to their specific needs and goals. For example, if a company deals with sensitive personal information such as health records or location data, it should prioritize security aspects above all else.
My recommendation to data leadership teams:
1. Manage data teams as profit centers: Prioritize profit-driving items like developing new products, reducing customer acquisition costs, or increasing conversion rates. These elements can be measured in a quantifiable way and their value can be easily demonstrated to stakeholders.
2. Build an incremental and iterative roadmap and validate assumptions at each stage: This approach allows businesses to prove their value using KPIs and analytics before moving on to more advanced machine learning (ML) and generative AI applications.
3. Don’t spend too much on infrastructure: Avoid managing data teams as pure infrastructure or technical teams, as this can lead to being viewed as a cost center. Again follow a gradual approach.
To effectively calculate ROI, businesses must take a holistic approach that goes beyond simple financial metrics. Data initiatives require a cultural shift within an organization to use data effectively, and this change must be supported by strong leadership and a clear vision.
By taking a thoughtful and strategic approach to their data initiatives, businesses can effectively calculate the ROI of their investments and achieve their business goals.
