In today’s business environment, it’s survival of the fittest. Expectations from customers and business partners have changed, making modernization even more critical. Yet change is notoriously difficult. As companies strive to be data-driven, they recognize that the transformation requires a new way of operating. It’s not enough to paint racing stripes on a VW Beetle to make it faster and more efficient. Same for companies. It’s a whole engine overhaul.
Data has traditionally been stored in silos, often associated with enterprise applications, maintained by separate teams, protected by specific functions, and residing in different locations across different cloud providers. These silos aren’t just within an organization’s four walls. They extend to external partner and supplier data. And, they tend to proliferate with mergers and acquisitions.
Maintaining these data silos is expensive and time-consuming; governance and collaboration are nearly impossible across multiple technologies and clouds. And, siloed data isn’t easy to find and use, and even more difficult to derive value from.
Yet there is hope. And, that hope starts with thinking about how we use the data. Imagine a customer 360, which pulls data from various sources or touchpoints with the customer: a CRM application, sales transactions, web logs, product usage, all enriched with partner or other third-party information. This 360-degree view enables a better understanding of customers, which enables product teams to improve offerings, marketing teams to personalize outreach, and customer service teams to improve customer experience.
Similarly, we could create a product 360, pulling data from various production processes or software embedded in them, plus sales transactions, service records, and product usage. This 360-degree view of the product enables the relevant teams to predict maintenance, reduce downtime, detect anomalies, and ultimately improve product development, sales, and customer experience.
These examples are data products, the new building block powering the engine of a data-driven company. And that message seems to resonate with data leaders. According to the CDO Agenda 2023, 38% of CDOs have adopted data product management with product managers.
A data product is a reusable building block, built to deliver data, or insights from that data, for a specific purpose. Thinking of data in this way requires a new way of thinking and organizing teams, as both producers and consumers need to reconfigure their teams to adapt to data product thinking. Data producers deliver data products from a single source or set of sources, such as data from a CRM application. Those data products could be used by themselves or aggregated into an aggregate data product, like the customer 360 described above. The consumers of the customer 360 must be able to articulate their requirements for that aggregate data product.
Product thinking works from the outside in. Think of it as “working backwards.” The process starts with asking how the data would be used and by whom. What are they trying to achieve and what do they need to achieve it? These customers and use cases could be internal or external. Here are a few examples:
These scenarios suggest different types of data products. Knowing customers requires collecting different attributes and creating personas or segmenting audiences. Prioritizing requires ranking. Predicting requires analytic models. So after thinking about the audience and the use case, you’d need to design the right data product for them.
A data product can take different forms, like any type of product. Imagine a data table, a report or dashboard, an analytic model, or a decision-support tool that enables a decision-maker to “play” with the data. During the panel I moderated at the Snowflake Summit last June, Amin Venjara, GM of Data Solutions at ADP, described the choices as “sand, glass, or jewelry.” I liked the analogy but I modified it slightly.
Do you offer sand, the raw material, to someone who will do something with it? Sand needs to be turned into something else. Do you make a glass, something that can be used by someone to deliver value? A glass is an intermediate step. It serves a purpose, but is only really valuable when combined with something else like water or wine. Or do you create something more refined, a unique product for a specific audience and purpose? The lamp is designed for a specific purpose. It illuminates a room, and delivers an immediate value.
Apply this analogy to data products. Sand is the raw material, the data. If you are a data producer and you need to deliver a product to a data scientist or developer, you’d likely deliver clean, ready-to-use data. Many data providers have traditionally offered raw data. Despite other options, when considering data products, demand for raw data remains. At last year’s Snowflake Summit, a post-trade financial services company, described some of the buyers of their data products as highly sophisticated hedge fund analysts. They don’t need a pre-built model. They want the raw material, and have the skill to build something with it.
The glass is a product of the data but not the final product. It delivers value but only when used in combination with something else. This is a data service like identity resolution or data enrichment. IPinfo, for example, provides location data enrichment for IP addresses to help companies know more about the origin of internet traffic, assigning geolocation-based attributes to internet traffic, monitoring and controlling website access by country, compliance with data protection regulations, digital rights management, fraud detection and prevention, and more.
The lamp is purpose-built, an application designed for a specific use. This is an application based on underlying data that could help a decision-maker streamline a process, optimize prices, personalize marketing campaigns, or optimize marketing mix. Or the application could be embedded, triggering an event or action based on streaming data inputs.
Data applications are products built with a specific use in mind. This could be data from a single source or multiple sources aggregated to address a specific business use case. Snowflake’s Product Manager’s Guide to Building Data Apps on the Data Cloud provides examples of common use cases such as a customer 360 that aggregates data from multiple sources to get a better understanding of customers and prospects. But even the customer 360 application could be tailored to specific audiences across sales, marketing, or customer service to be used to create customer segmentation, lead prioritization, personalized offers or experiences, and better customer service.
Several great examples of data products come from Scania, a Swedish manufacturer of heavy trucks, buses, and engines. Scania’s Connected Analytics Platform ingests approximately 150 million streaming messages per day from a fleet of 600,000 connected vehicles. Those messages combined with other data are important to different teams within the company and its customers. For example, Scania’s Telecom Analytics data product analyzes massive amounts of telecommunications data to ensure optimal connectivity. Comparing costs and revenue data for connected vehicles helped Scania optimize subscription profitability. Scania’s Digital Services Intelligence data product makes it easier to track and recommend subscriptions for each customer. Scania also uses connected vehicle data to inform pricing decisions for services.
The data products also benefit external customers. Scania’s predictive models recommend maintenance based on vehicle operation and workshop availability. The aim is to eliminate downtime for the end customer. A newer product, Scania’s BEV Analytics data product, provides a baseline for understanding electric vehicle range, battery life and performance, and total cost of ownership. “Interest in this data product is huge,” says Ulf Holmström, Lead Data Scientist at Scania. “BEVs [battery electric vehicles] are the future, and everyone wants to see how they operate across different dimensions.” As the electric vehicle market grows, interest in this data product could go well beyond existing customers and partners.
As demand for third-party data grows so do the opportunities to build new revenue streams from data products. A first step is to offer data products or services as an incremental offering, such as insights into product usage or benchmarking costs, or revenues against others in a similar industry or region. On top of an existing product or service, this incremental offering is like asking, “Do you want fries with that?”
For example, as a provider of HR and payroll solutions, ADP processes almost 20% of U.S. payrolls, capturing vast amounts of near real-time data on salaries, skills, and benefits. It knows how much companies are spending on employees, who’s hiring, who’s firing, and what that looks like across different regions.
As a first step, ADP began offering insights to its customers. How does your retention rate compare to others in your industry? Or in your region? Are your salaries and benefits competitive? With machine learning applications across its HR, Payroll, and time applications, customers can also optimize demand planning.
Insights and trends from this data also provide a view of the U.S. economy that’s more granular and timely than data from the U.S. Department of Labor Statistics. ADP launched a National Employment Report and licenses the data to help companies with site selection, for example, based on local skill sets and salary expectations. With these new products, ADP has become a data provider both to existing customers and to net new customers for a whole new revenue stream.
For more on the data product journey—from identifying audiences, use cases, and the form of the data product to pricing options and choosing which channels to market—please join the session on Data Commercialization: Your Guide to Taking Data to Market at the Snowflake Summit 2023. You’ll also find plenty of Summit sessions on data apps. Check out Build an App for That: The Next Big Opportunity for Data Entrepreneurs with Mode CTO Benn Stancil, or get a little more technical in the Build Your Snowpark-Powered Data Products and Data Applications with DataOps.live session.
The post <strong>Are You Data Economy Ready? Start with Data Product Thinking</strong> appeared first on Snowflake.
We had a busy week catching up with customers and partners at Microsoft Ignite in Chicago and online. We shared […]
At Snowflake, we’re committed to delivering consistent, automatic performance enhancements. We work behind the scenes to make your data operations […]
For years, companies have operated under the prevailing notion that AI is reserved only for the corporate giants — the […]
Our services help both the big and small organisations in the identification of security breaches and provide outclass solutions to prevent any potential harm.