Usage-Based Pricing
people reading graphs
More and more businesses are leveraging usage-based pricing (UBP) to drive customer and revenue growth. Learn more here.
Volume XXVI, Issue 19 |

A recent L.E.K. Consulting piece highlighted the growing appeal of usage-based pricing models for small and medium-sized software-as-a-service (SaaS) providers. As software purchasing becomes more decentralized across business units, buyers have gained leverage to demand pricing tied directly to product value and outcomes.

However, enterprise customers tend to prioritize pricing/budget predictability over flexibility. This creates a conflict between pricing to value (and/or utilizing a consistent pricing model and metric(s) across customer segments) and addressing customer needs.

This edition of Executive Insights explores strategies to balance usage-based approaches with the financial control that large organizations prefer, detailing available models for enterprise SaaS businesses to meet varying customer needs.

Usage-based pricing models

In our usage-based pricing piece, we discussed how pay-as-you-go (PAYG) pricing is growing in popularity among SaaS companies to drive revenue growth by charging based on customer usage rather than rigid subscription fees. PAYG helps lower barriers to entry, enables incremental upsells as usage grows and better aligns value to price (see Figure 1).

Source: L.E.K. research and analysis

PAYG models first became prominent because of their use by cloud infrastructure services. Trends like shifting buying power to customers and decentralized software purchasing have further accelerated PAYG adoption. Over 60% of SaaS companies now leverage PAYG to some degree. However, basic PAYG presents challenges, especially for enterprise customers that value pricing predictability.  

Let’s instead explore a range of models that increase predictability for enterprise SaaS customers.

Burstable reserve model

This model determines a baseline level of usage (e.g., per hour, per day, per week) but allows temporary resource scaling (“bursting”) during demand spikes. This is a particularly common feature in cloud services.

Microsoft Azure exemplifies this model with a pool of accumulating central processing unit credits. Credits build up during low-usage periods, funding bursts above baseline capacity in periods of higher utilization (see Figure 2). This ultimately smooths spend across usage peaks and troughs. 

Source: Ampere

Total cost = Base compute + storage + overage credits for spikes

Variations of this model can add further spend predictability around the concept of burstable reserve. For example, a basic tier may only allow the usage of “overage credits” that have already been accrued (i.e., during prior periods of below-baseline utilization), while a more advanced tier may allow for the usage of credits that have not yet been accrued (i.e., “borrowing” from future periods). This is one example of how usage-based pricing can be combined with more traditional pricing techniques, such as price tiering.

Adaptive flat/volumetric usage-based pricing

In this model, customers prepurchase a set allowance of “usage units” based on their anticipated usage (e.g., compute, credits/tokens, application programming interface (API) calls) (see Figure 3). In enterprise pricing, this usage band usually covers a year of usage. If usage surpasses the initial allowance, customers transition to a higher band with new corresponding rates.

Source: Bessemer Venture Partners

In this hypothetical volumetric pricing model, the rate is $0.06 per API call if prepurchasing within an allotted volume. Overage fees of $0.07 per API call apply when usage goes above the allotment. The formula is:

Total cost = (Base rate x prepurchased credits) + (higher band rate x overage)

Greater or lesser predictability is driven by the design of the unit bands:

  • Broad bands: These provide greater certainty for customers, reducing the risk of unexpected costs, but have more limited upside for vendors
  • Narrow bands: These increasingly approach a PAYG model, requiring more accurate usage predictions from customers but offering finer control over costs and usage

Adaptive flat/volumetric usage-based models can also be combined with more traditional pricing techniques, such as offering savings for bigger commitments and monetizing overages at a higher rate.  

This pricing model is tailored to align customer needs for predictable expenses with the provider's goal of maximizing revenue opportunities, balancing flexibility and predictability.

Annual commit-and-drawdown model

In this pricing model, customers make an upfront commitment to a specific volume of credit units but have the flexibility to adjust based on the actual usage over the contract period.  

Depending on the size and nature of the agreement, this model may require robust customer communication on usage levels and potential true-up fees — for example, notifying customers midyear if over 50% of credits have been used. This transparency into drawdown and projected end-of-term charges is essential.  

The annual true-up ultimately aligns consumption to upfront commitments, charging for overages or crediting remaining allotments. This blend of flexibility and predictable spending aligns customer and provider incentives.

Other considerations: Tailoring the approach

Rather than a one-size-fits-all approach, vendors should tailor their usage-based pricing strategy while considering several critical dimensions:

Use cases: Some applications have predictable seasonal spikes such as financial period closing activities. Others see variability around events like new product launches or M&A. Models can flex to adapt to known use cases through accrual and true-up mechanisms.

Number and types of users: Within a single enterprise, different groups — power users, knowledge workers, third-party partners — have distinct usage patterns. Units and allotments can be tiered based on each group’s expected consumption.

Customer industry and size: Credit unit needs will vary significantly across verticals like financial services, healthcare, retail, etc. Small, midsize and large enterprise needs also differ. Models should be adapted accordingly, with appropriate baselines and bands.

Vendors should gather data on these dimensions and design tailored pricing bands, credit pools and overage calculations. This reduces uncertainty for customers while maximizing potential value.  

Further insights

  1. Overage models: How should businesses charge customers when usage exceeds initially agreed-upon credit limits under different pricing models? We will explore approaches ranging from PAYG overflow pricing to capped overage pools.
  2. Billing strategies: Should revenue be recognized through recurring monthly invoices or an annual true-up? A meaningful difference exists between how customer usage accumulates and how actual revenue is captured under flexible-pricing approaches.

As your business evaluates usage-based models, consider that we have decades of experience helping clients implement adaptive approaches to optimize value.

L.E.K. Consulting is a registered trademark of L.E.K. Consulting LLC. All other products and brands mentioned in this document are properties of their respective owners. © 2024 L.E.K. Consulting LLC

Questions about our latest thinking?

Related Insights