November 14, 2023 By Jeremy Caine 7 min read

Data monetization is a business capability where an organization can create and realize value from data and artificial intelligence (AI) assets. A value exchange system built on data products can drive business growth for your organization and gain competitive advantage. This growth could be internal cost effectiveness, stronger risk compliance, increasing the economic value of a partner ecosystem, or through new revenue streams. Advanced data management software and generative AI can accelerate the creation of a platform capability for scalable delivery of enterprise ready data and AI products.

Why data monetization matters

According to McKinsey in the Harvard Business Review, a single data product at a national US bank feeds 60 use cases in business applications, which eliminated $40M in losses and generates $60M incremental revenue annually. In the public sector, Transport for London provides free and open data across 80 data feeds powering over 600 apps and contribute up to £130m to London’s economy.

Data monetization is not narrowly “selling data sets;” it is about improving work and enhancing business performance by better-using data. Internal data monetization initiatives measure improvement in process design, task guidance and optimization of data used in the organization’s product or service offerings. External monetization opportunities enable different types of data in different formats to be information assets that can be sold or have their value recorded when used.

Creating value from data involves taking some action on the data. Realizing that value is the activity that ensures there is an economic benefit from the created value that contributes to the organization’s bottom line.

Data monetization strategy: Managing data as a product

Every organization has the potential to monetize their data; for many organizations, it is an untapped resource for new capabilities. Data-as-a-Service and data marketplaces are well established to create data value from initiatives built on data analytics, big data and business intelligence. But few organizations have made the strategic shift to managing “data as a product.” This data management means applying product development practices to data. High performing, data-driven organizations have created new business models, utility partnerships and enhanced existing offerings from data monetization that contributes more than 20% to the company’s profitability.

The key play is to treat data as a strategic asset with a user-centric product approach where this new product can be consumed by a diverse set of applications. Organizations build trust in their data and AI by demonstrating transparency and ethics, recognizing data privacy, adhering to regulations, and keeping data safe and secure.

Data products and data mesh

Data products are assembled data from sources that can serve a set of functional needs that can be packaged into a consumable unit. Each data product has its own lifecycle environment where its data and AI assets are managed in their product-specific data lakehouse. Flexibility in data collection is made possible when product lakehouses connect to and ingest data from many sources, using many different technology protocols. Additionally, by managing the data product as an isolated unit it can have location flexibility and portability — private or public cloud — depending on the established sensitivity and privacy controls for the data. IBM offers connectivity flexibility and hosting of data product lakehouses built on Red Hat OpenShift for an open hybrid cloud deployment.

Get started with

Data mesh architectures have now emerged as the cost-effective way to serve data products to a variety of endpoint types, with detailed and robust usage tracking, risk and compliance measurements, and security. One or more data products are served over the mesh and consumed by an end-user application as an auditable transaction.

For example, a financial markets business might offer one product giving a real-time market data feed and another offering finance-related news. A consumer might build a decision-making application leveraging both of those products and offer trading options based on price and political or environmental news.

Building a solution capability for data management

Developing a capability depends on the ability to “connect the dots” for your stakeholders. It is a supply chain from your raw data sources to the transparent and traceable exchange of value when a data asset is consumed in an end-user experience.

You can do this by developing a solution framework for data monetization that incorporates:

Three stages of the data monetization lifecycle process:

  1. Create: Raw data is accessed and assembled into data products.
  2. Serve: Data products are discoverable and consumed as services, typically via a platform.
  3. Realize: Each data service has an agreed unit of value which is transacted and measured.
Figure 1: The Data Monetization Lifecycle

The raw data that fuels data monetization will come from three source categories: enterprise systems, external data and personal data. Data products are assembled from operational and analytical views of company and customer data which can be combined with public data sets (not necessarily free). Personal data is an important view across enterprise and public data that needs to be managed correctly by an organization. When a “right to be forgotten” request is invoked it spans from the raw data source to the data product target.

Data products come in many forms including datasets, programs and AI models. They are packaged and deployed for consumption as a service, and there can be multiple service types for any one product. Popular service consumption types include download, API and streaming.

Take the example of a client who integrated a set of disparate company ESG data into a new dataset. Their data services were a full dataset download plus an API wrap around the data, which could be queried for ESG data based on a company ticker symbol.

Data monetization is about realizing value from data. The consumer of data product services needs the ability to find and assess a product, pay for it and then invoke one or more of the service endpoints to consume it. Depending on the user’s business model they may be consuming that service for their own use in their capabilities, or under appropriate license terms to create a downstream product or customer experience using the data product for their own revenue stream.

Figure 2: The Data Monetization Value Chain

Achieve scale with a platform approach

A user’s options for consumption vary. The traditional approach may have been for the data product supplier to distribute its large one-size-fits-all datasets direct to clients or into multiple data marketplaces. For APIs, they may have built them into the catalog ecosystems of each hyperscaler cloud catalog. For AI models and associated datasets, they could look to utilize a marketplace like Hugging Face. These consumption provisions now start to become a complex, federated and less cost-effective way of maximizing profit from transaction and subscription revenues.

With the data monetization solution framework, the maximized return on value can come from a data product provider becoming a data SaaS business. The framework defines a reference architecture that integrates a set of technologies and products including IBM Data and AI products.

Figure 3: Implementing the Solution Stack with IBM Data and AI

Implementation across the full lifecycle covers:

  • Create: Ingest source data sets and feeds and transform these into data product assets using hybrid cloud lakehouse technology with integrated data science and AI development environments.
  • Serve: Build cloud services for data products through automation and platform service technology so they can be operated securely at global scale.
  • Realize: Instrument the data product services to enable adherence to risk and compliance controls with event and metrics data integrated to financial management.

A further extension on this SaaS capability is where the data product provider also offers a multi-tenant, multi-user creator environment. Multiple parties collaborate in their own development spaces, consuming the data product services on the platform in their offerings and then hosting for consumption by their customers.

Enterprise artificial intelligence

Many organizations have built mature software systems with machine learning and deep learning functions to power their business processes and customer offerings. Generative AI has only served to accelerate the options for data product design, lifecycle delivery and operational management.

Platform builders and operators can use AI models to build tools. Creators can use those tools to discover or learn about data in enterprise systems and public domain. Code generation “co-pilot” tools (e.g., watsonx Code Assistant) build and maintain automations and create natural language driven experiences for operations or customer service. These add to established practices of using AIOps and advanced analytics around finance and risk functions.

Data product owners and service owners can innovate with Generative AI tools. They can augment data set assembly with generated synthetic data and create new analyses of data sources, which in turn can eliminate outliers and anomalies. Doing so can increase the quality of data integrated into data products. It can be used to develop data product specific classification and knowledge bases of data sets, as well as build organization and domain specific AI models to offer as products.

Enterprise Generative AI is beginning to orient itself around what are the right type of models and training approaches. More importantly they are looking at the trust and transparency of the datasets these models are trained on, plus the legal indemnification position when using them.

Data product owners that are building or integrating such models must consider trust and transparency when designing the value exchange. By utilising, an organization’s data monetization roadmap can take advantage of models such as IBM Granite to be assured of its transparency and indemnification.

Accelerating data monetization

The foundational products that can be used to build the platform are IBM Cloud Pak for Data and IBM Cloud Pak for Integration. Their components enable development of data products and services for end user consumption at production scale. adds data product lakehouse abilities and adds advanced generative AI model development.

A cohesive data monetization service experience can be built with these products and pre-designed automation workflows. Built to run on Red Hat OpenShift this gives the advantage of a scalable platform that can be built once, deployed across multiple private on-premises and public cloud infrastructures, and run as a consistent single service.

Using this IBM solution framework organizations can shift to using data as a strategic asset and inject innovation into the business model through enterprise AI technology.

Explore enterprise generative AI with watsonx technology
Was this article helpful?

More from Artificial intelligence

IBM watsonx Challenge empowers partners to solve real-world problems with AI

2 min read - In June, IBM invited ecosystem partners in Europe, the Middle East and Africa to participate in an IBM watsonx™ Challenge, a hands-on experience designed to bring the watsonx platform capabilities to some of the most important members of the IBM ecosystem. These ecosystem partners, who sell, build or service IBM technologies, enthusiastically embraced the challenge. Participants formed teams and focused on quickly crafting a solution to one of three selected challenges.   The challenges included using prompt engineering to analyze…

10 tasks I wish AI could perform for financial planning and analysis professionals

4 min read - It’s no secret that artificial intelligence (AI) transforms the way we work in financial planning and analysis (FP&A). It is already happening to a degree, but we could easily dream of many more things that AI could do for us. Most FP&A professionals are consumed with manual work that detracts from their ability to add value to their work. This often leaves chief financial officers and business leaders frustrated with the return on investment from their FP&A team. However, AI…

ServiceNow and IBM revolutionize talent development with AI

4 min read - Generative AI is fundamentally changing the world of work by redefining the skills and jobs needed for the future. In fact, recent research from ServiceNow and Pearson found that an additional 1.76 million tech workers will be needed by 2028 in the US alone.  However, according to the IBM Institute for Business Value, less than half of CEOs surveyed (44%) have assessed the potential impact of generative AI on their workforces. To help customers develop and upskill their workforces to meet…

IBM Newsletters

Get our newsletters and topic updates that deliver the latest thought leadership and insights on emerging trends.
Subscribe now More newsletters