The Evolution of Pricing Models for SaaS Companies

By John Pineda, David Lin, Nikhil Singhal, and Stella Su

A decade ago, subscriptions were the dominant model of software as a service (SaaS) pricing. At first, subscriptions were based on the number of users, seats, or licenses, packaged in monthly or annual contracts. Then, as SaaS companies expanded and evolved their products, some started pricing subscriptions based on other value-aligned metrics (such as devices, sites, etc.). Recently, however, pricing has undergone a different sort of shift: from subscriptions to usage. In usage-based pricing, customers pay only for what they use (for example, size of data stored on cloud, number of transactions run, etc.). As BCG’s IT Spending Pulse surveys from 2018 to 2023 show, IT buyers have started to shift toward hybrid (combination of subscription and usage pricing) and pure usage-based pricing models.

While interest in usage pricing is growing in all types of software stacks, infrastructure as a service (IaaS) and platform as a service (PaaS) are adopting it at faster rates. For example, most major cloud service providers today (cloud computing, cloud storage, networking, databases) utilize a usage-based pricing model in which customers are only charged for the amount of cloud resources they use. Companies that adopt usage pricing also outperform their peers and show more resilience in valuations, as their pricing is able to scale with the usage of their products.

The shift toward usage pricing is further accelerated by the popularity of OpenAI’s GPT and other Large Language Models that charge using “tokens.” As companies across verticals begin incorporating GenAI into their own products, they will likely look to pass down the usage cost. A recent BCG survey finds that a majority of respondents prefer a usage-based or revenue/value share pricing model when engaging with GenAI-based services (see Exhibit 2).

A recent research study published in partnership with Zuora dives into the growth and benefits of consumption pricing, and also offers guidance on the “what,” “when,” and “how” of consumption pricing — whether it be usage-based overages on top of subscriptions, usage-based pay-as-you-go, or some other types of hybrid models in between. You can read more here.

The following explores the spectrum of SaaS pricing models and how the market has evolved beyond pure usage-based pricing into what we call “hybrid” models.

Archetypes of Pricing Models

As the spectrum in Exhibit 3 above demonstrates, in general, pricing models range in terms of time to value for vendors or customers. On the far left are “perpetual” models that charge upfront, one time fees, generating immediate payment for vendors but generally being less aligned with customer value, as the customer does not benefit from software upgrades. On the far right are “outcome-based” models in which vendors are only paid if their software achieves certain outcomes for customers, delaying payment for vendors but being directly aligned with customer value.

Both subscription and usage-based models fall in the middle of the spectrum. To break these two categories down further, see this “zoomed in” spectrum shown in Exhibit 4.

On the left-hand side of the spectrum are “classic subscription” models wherein customers pay a recurring fee (usually monthly, semi-annually, or annually) for access to the product.

On the right-hand side of the spectrum are “usage-based” models in which customers pay as they go only for their actual usage.

The key difference between pure usage-based pricing and pure subscriptions is direct commitment to usage of the product. For subscriptions, customers pay upfront for unlimited access to the product; for usage-based models, they pay only their actual amount of usage.

The middle column refers to the hybrid pricing models in which companies combine the benefits of both subscription and usage models. The following are examples of different flavors:

New Relic is a great example of how a hybrid pricing model works in practice. A look at New Relic’s pricing page (see Exhibit 6) shows how the company charges on two different axes: data costs and user costs.

User costs are monthly, recurring fixed costs depending on the number of users/seats purchased — a classic subscription model. Data costs then sit on top of subscription costs, where a customer’s monthly cost for data depends on the amount, and also type, of data used. In New Relic’s case, a company will pay $0.30/GB for normal data usage, and $0.50/GB for Data Plus usage (beyond the free 100 GB). From a vendor perspective, this generates predictable, recurring revenue (user costs) and also captures variable, consumption-driven revenue (data costs).

Benefits of and Challenges to Hybrid Pricing

Finding one’s sweet spot on the spectrum of options is all the more important as a SaaS company grows and its product portfolio becomes more sophisticated and complex. While straight-forward pricing models may have been effective in the past for SaaS players, there is potential money left on the table if pricing does not evolve with their products.

There are clear benefits from taking a hybrid subscription and usage approach in which customers leverage aspects of both pricing approaches, including the following:

• Faster adoption and easier scalability of products as well as identification of new use-cases within an organization, as higher usage rates do not need to be re-negotiated
• Better alignment between pricing and value, as buyers are only paying for as much of the product as they are using
• More predictability, as pricing is fairly constant at the start, which is attractive to both vendors and customers. If customers have a good sense of what their usage will be, they will generally pay for what they anticipate using, which creates predictability in their budget.
• Improved revenue capture, as customers pay for the entire cost of their pre-committed tier even if their usage is below the tier, thus improving revenue capture
• Encourages usage, as customers are also incentivized to maximize usage within their tier, thereby increasing the value they receive from the products and raising switching costs
• Targeted packaging, which can capture different willingness to pay across customer segments. For example, a simple “Good- Better-Best” packaging can create guardrails on the volume of usage as well as the complexity of features and inclusion of additional services (for example, platform management, technical support, analytics, etc). Companies can design smart tiers for different customer segments; one segment may want high volume usage but simple features, another segment may want complex, high-value features with additional services but will have low usage, etc.

Despite these benefits, there are potential challenges to consider, including the following:

• Customers that want more predictable pricing may prefer simple subscription pricing even over hybrid models.
• Customers may experience frustration with the need to track usage, especially if metering is variable and across multiple products. From the vendor perspective, providing real-time visibility for usage to customers is crucial, especially if they are charging for overages.
• While scaling product adoption may be theoretically easier with usage pricing, it can also work the other way, as organizations may feel incentivized to limit usage because they have to pay every time they use a product. As a result, adoption and experimentation may be disincentivized.
• With GenAI, for instance, a pure usage model may frustrate customers, as it may take multiple iterations before they can generate something of value.

Beyond customer behavior, there are secondary costs to usage pricing that are not always taken into account. Pricing based on any sort of usage metric invites significant variability to realized revenues, which is not welcomed by the market and may, in turn, affect valuations or lead to stock price volatility. Also, while usage metrics are fairly easy to identify, they still need to be tracked and may create back-end complexity and billing issues and thus lead to customer disputes.

In conclusion, there is not a “one-size-fits-all” solution for all companies. There needs to be careful consideration of what model works best for a company’s context. Pure usage or hybrid models offer a myriad of benefits and flexibility; however, they also add more complexity to pricing operations and sales cycles. Introducing a “usage” or “pricing” calculator on a company’s internal website and training sales representatives to deeply understand the needs and rhythms of their customers are just two examples of what it might take to successfully roll out a hybrid pricing model. As SaaS companies evolve their pricing models to fit market trends and customer demands, they also need to find a way to contain the complexity.