A report by Bloomberg Intelligence projects the AI industry will reach $1.3 trillion by 2032, with proprietary data fueling much of this growth. As businesses increasingly adopt generative AI (genAI) to enhance efficiency, data is rapidly becoming one of the most valuable assets in the digital economy.

Foundational AI models require vast amounts of data for training, and many AI products are now leveraging proprietary datasets alongside these models to power innovative applications and AI agents. These tools have the potential to transform business processes across engineering, sales, support, and beyond.

Chances are, your organization already holds a wealth of proprietary data. Whether it’s internal data supporting a traditional SaaS application or user-generated content, this data isn’t just a strategic asset for internal use — it can also be monetized by selling it to enterprises that need high-quality datasets for training their models or driving their applications.

Monetizing data can bring numerous benefits to an organization. By leveraging their existing data, companies can unlock new revenue streams and gain a competitive edge in the market. But how do you unlock this value while navigating the challenges of monetizing your proprietary data?

Understanding Data Monetization

As a data provider, balancing customer needs with revenue growth is no small feat, and without a well-thought-out data monetization strategy, you risk losing potential revenue, stifling your growth, and limiting your ability to scale. Imagine a worst-case scenario: A customer signs up, downloads all the data they need in a single day, and never returns. Effective data monetization strategies can mitigate these risks by aligning pricing models with customer usage patterns and ensuring sustainable revenue growth.

Unlike traditional API businesses, where API consumption is typically predictable “up and to the right,” data consumption often follows a sporadic pattern. Customers typically consume data only when needed. For example, if you’re providing data to assist marketing teams, they might only need the data ahead of large marketing launches. Similarly, if you are providing financial data around real estate transactions, you may find customers only care about the data during end-of-year planning or ahead of the spring buying season.

Simply charging a flat monthly or yearly fee might not align with the value customers receive, especially when their consumption is irregular. This raises an important question: How can you ensure predictable revenue and cash flow for your business while reducing obstacles for customers whose usage fluctuates and is unpredictable?

Different API Monetization Models

One effective approach to monetizing data is usage-based billing (also referred to as consumption-based billing). This model allows customers to pay only for what they use, offering flexibility and avoiding the commitment of a subscription. Moreover, it enables your revenue to scale naturally as customers’ data needs grow.

Cloud providers and API platforms have widely adopted usage-based billing. It’s in practice in both modern SaaS companies like NexHealth and traditional enterprises like Siemens. A typical implementation involves tracking API usage over a billing period (such as a month) and invoicing customers at the end of that period. This model works well if the cost of providing the API is low and the risk of abuse is minimal.

However, data providers often face higher costs of goods sold (COGS) or risks of misuse. For example, a customer might download all the data they need and then cancel their subscription or simply fail to pay their invoice. To mitigate these risks, many providers are adopting a prepaid pay-as-you-go (PAYG) model.

Modern AI companies like You.com and OpenAI, along with telcos such as Sinch and Twilio, are leveraging PAYG to help align usage-based revenue to their usage-based cost. With prepaid PAYG, customers purchase credits upfront, which are then consumed based on a pre-negotiated rate — similar to buying a prepaid phone card. This model reduces the risk of abuse and provides immediate cash flow for your business, making it a win-win for providers and customers.

How to Meter API Data Consumption

Even with a PAYG model, determining how to charge customers requires careful consideration. Metering by API calls alone is often ineffective, as customers prioritize efficient batch queries to maximize throughput. A single API call could result in the export of massive datasets.

For example, if you are offering a financial data enrichment API, your customers may want to enrich thousands or millions of records in a large batch. In this case, the ideal flow would be that the customer submits a batch job for all the entities needing enrichment. Since this is a batch job, some items may not be found or fail. Customers shouldn’t be charged for missing or low-quality data.

To address this, it’s important to align your billable consumption metrics with customer value. For example, you could charge per successful data element or row accessed, excluding rows that are incomplete or of poor quality. Tracking and metering such granular data usage can be complex and typically requires additional monitoring tools to analyze API consumption effectively.

Managing Asynchronous Jobs

For APIs that involve backend jobs (like exporting large datasets), asynchronous processing adds another layer of complexity to monetization. You must decide when to deduct credits from a customer’s balance and how to handle failure scenarios.

A common approach is to “lock” the credits until the job completes, ensuring customers cannot trigger excessive jobs that would cause their balance to drop into negative territory.

Job Handling Scenarios:

Example: Partial Completion

1. Customer makes an API call to fetch 1,000 items.
2. Customer balance reduced by $1000.
3. Backend job triggered to export 1,000 items.
4. Job completes, but 100 items are missing.
5. Customer credited back $100 for the missing items.

Example: Failure Case

1. Customer makes an API call to fetch 1,000 items.
2. Customer balance reduced by $1000.
3. Backend job triggered to export 1,000 items.
4. Job fails.
5. Customer notified and credited back the entire $1,000 due to job failure.

Measuring Data Quality and API Experience

Providing high-quality data through a seamless API experience is essential for retaining customers. Leveraging data analytics can help you measure and improve this quality.

Unlike traditional APIs, where success can be measured by HTTP response codes (like 200 vs. 400), evaluating data quality is far more nuanced. A common approach is to apply a different number of credits depending on the quality.

A recommended approach is to assign a Response Quality Score using the formula:

Response Quality Score = sum( accuracy of row * relevancy of row ) / total rows

This scoring system helps developers understand where their API descriptions may fall short and how they can improve alignment.

Security and Regulatory Considerations

Selling proprietary data via APIs comes with regulatory and security responsibilities, especially when offering data as a service (DaaS) through the cloud. Before diving in, ensure you have the legal right to sell the data. Some datasets may be governed by copyright laws or regulations like HIPAA, PCI, or GDPR. For example, GDPR’s “right to be forgotten” requires mechanisms to delete specific data upon request.

Additionally, it’s critical to implement robust security measures to protect sensitive information and build customer trust. This includes encrypting data with either server-side encryption or client-side encryption, securing API endpoints, and maintaining compliance with relevant data privacy standards.

The Future of Data Monetization in the AI World

In an era where generative AI drives innovation across industries, proprietary data has become an increasingly valuable resource. In fact, data marketplaces are becoming crucial platforms for buying and selling data. By monetizing your data, you can transform it from a cost center into a profit center, unlocking new revenue streams while fueling advancements in AI.

With the right monetization model and a focus on delivering high-quality, valuable data, you can position your organization at the forefront of this rapidly evolving landscape.