Software monetization in the AI era

AI features demand a shift in monetization

Two dominant models have emerged as industry standards for software monetization and pricing. The first is a good-better-best price model tailored to customer needs ("jobs to be done") and providing clear upsell paths. The second is a user-based price metric that aligns with the customer's ability to scale their business and, consequently, their use of the software.

In a world where service costs are only minimally linked to actual software usage, this model has proven effective for software players, as profits scale perfectly with the increase in the number of users. Leading US companies, such as Microsoft and Salesforce, pioneered the shift from a license-based, one-off model to a user-based subscription model, marking a turning point toward annual recurring revenue (ARR) or monthly recurring revenue (MRR) as key valuation metrics.

Over the years, this model has gained widespread application across the industry and evolved further, with players like Slack introducing variations such as "price per active user."

However, with the rise of AI and generative AI (GenAI), particularly the recent large language models (LLMs) powering OpenAI's GPT, Google's Gemini, and Meta's Llama, a shift in monetization strategies is visible and necessary. Unlike traditional SaaS models, the degree of utilization of (Gen)AI features directly impacts the costs for the software provider—the more AI tasks are processed, the higher the costs for the provider. While advancements in LLMs and data centers will reduce costs, the financial model diverges significantly from that of software applications without AI.

Insane thing: we are currently losing money on OpenAI Pro subscriptions! People use it much more than we expected.

OpenAI CEO Sam Altman's recent post on X

This post came just a few weeks after OpenAI launched its new ChatGPT Pro subscription for $200 monthly.

New principles in packaging and pricing

AI is compelling software providers to rethink their monetization strategies. As AI-powered features become embedded in products, companies must revisit both packaging and pricing models. Introducing AI into traditional SaaS offerings changes how those products are differentiated and, thus, how they should be packaged.

SaaS products have traditionally been differentiated by the features included in each package, with premium packages offering more advanced or numerous features to encourage clear upsell paths. With the integration of AI, this feature-based approach gains an additional dimension: AI capabilities.

Now, the service offered to the customer can be differentiated based on the level of AI functionality, with aspects of differentiation strongly oriented toward professional services. For example, factors such as accuracy and response time become key differentiators. Especially with the rise of agentic AI, packaging and pricing will increasingly resemble models used to price human expertise and skills.

Regarding the price model, adjustments to price metrics or the introduction of new price metrics are increasingly being applied. A shift from traditional SaaS metrics, such as user-based pricing, to consumption-based and outcome-based metrics is currently underway. Today, four principal pricing models are gaining traction:

• User-based pricing, also called per-user or seat-based pricing, is a model where the software subscription cost depends on the number of users ("seats"). Each user incurs a fixed fee and the total cost scales linearly with the number of users.
• Consumption-based pricing, also known as usage-based or pay-as-you-go pricing, charges customers based on their actual usage of a product or service.
• Outcome or output-based pricing links the cost of a product or service to the results or value it delivers. Customers pay based on the tangible benefits or quantitative output they receive.
• Hybrid pricing is an adaptive strategy that combines multiple pricing models, offering customers a flexible and tailored solution. It typically blends fixed-rate pricing with usage-based components.

Recent shifts in AI packaging and pricing strategies from Microsoft and Google illustrate different approaches to monetizing AI. Microsoft recently launched the new 365 Copilot Chat, which has a different pricing strategy compared to Copilot as a whole. Microsoft is charging for the AI feature using a consumption-based pricing model.

Google has taken a different path. It has decided not to charge extra for AI features in the corporate context. Instead, Gemini for Workspaces, which will include AI by default, will result in a $2 increase in the regular plan. These will be priced at a fixed amount for all users, equivalent to a user-based subscription pricing model.

Just recently, Salesforce has addressed the growing demand for digital work with major pricing updates for Agentforce. The new pricing model includes Flex Credits, where companies pay only for actions performed by Agentforce, costing 20 Flex Credits ($0.10) per action. The Flex Agreement allows companies to adjust investments based on business needs, providing flexibility to shift funds between user licenses and digital workforces as requirements change. Additionally, new Agentforce user licenses and add-ons offer unlimited use for employees at a straightforward monthly price. These innovations aim to facilitate AI adoption in businesses and increase investment flexibility.

Although the approaches of these three leading technology companies could not be more different, they share one common goal: to move away from high-priced add-on fees per user. The latest "2024 SaaS Benchmarks Report" by High Alpha provides a comprehensive overview of the current state of AI adoption among software companies and their exploration of new pricing models.
 

Figure 1: AI monetization

Five steps to successfully monetize AI features

1. Develop a new packaging structure. Begin by evaluating the capabilities of your software product and understanding the specific needs and requirements of your customers. This will help you define product differentiation and determine the potential market reach of the product, such as across different industry verticals. A well-thought-out packaging structure ensures that your AI features are communicated and aligned with customer expectations.
2. Evaluate potential price metrics. Consider various pricing metrics, such as user-based, consumption-based, and outcome-based models. Evaluate their impact on the customer's business, which ultimately translates to the costs for your product. It is crucial to ensure that the chosen price metric is operational, fits the customer's business model, is fair, and has strong monetization potential.
3. Define your price structure and differentiation. Combine your packaging options with a well-defined price structure. Decide how prices should vary based on factors such as region, customer size, or industry vertical. Determine how your pricing should scale and whether to incorporate multiple price metrics, such as a hybrid model that uses both user-based and consumption-based metrics. This approach allows for flexibility and better alignment with customer needs.
4. Test packaging and price model combinations. Conduct internal expert sessions, price experiments, customer surveys, and data predictions to test the derived combinations of packaging and price models. Evaluate the willingness-to-pay associated with each option. This step is essential for gathering insights and refining your approach based on real-world feedback and data.
5. Fine-tune and finalize. Based on the testing conducted, refine your concepts and determine the optimal packaging and pricing structure to monetize your product, including the AI features offered. This final step ensures that your offering is both competitive and aligned with market demands, maximizing the value delivered to customers and the revenue potential for your business.

Outlook

BearingPoint offers a comprehensive framework for assessing the monetization strategy of software players exploring AI opportunities. We support our clients at every stage, from creating new packaging and pricing options to testing them through market studies, customer and persona-driven interviews , and A/B tests.

Beyond strategy, we also assist in fine-tuning and implementing the new setup across the customer journey and internal processes. Our unique combination of consulting and software development capabilities, leveraged by our BE Products unit, makes us the ideal partner for software companies navigating the challenges of AI-driven change.