1. Choose a metric that you can link to customer value.

Subscriptions are all about value. Customers should be able to see the value they get from using your product. They should also be able to clearly tie that usage to their cost

2. Choose a metric that’s familiar. The right usage metric is also one that is transparent — in other words, customers clearly know what they’re paying for and can predict how a change in their usage will impact their billing. Transparency is key. Going with something esoteric won’t work. Whatever you choose should be both familiar to your customers and easy for you to track.

3. Choose a metric that your customers have reasonable control over.They should understand what drives their usage and be able to control it. This becomes even more important when you’re dealing with B2B customers.

Key question #1: Can customers see the clear connection between usage and value?

Key question #2: Is this metric obvious to your customers (i.e. something they are already aware of, and possibly even tracking)?

Key question #3: Do customers have control (and feel that sense of control) over this usage metric?

There isn’t just one way to “do” usage-based pricing, but there are some common approaches:

1. Per unit pricing.Per unit pricing (aka pay per use) is the most flexible and the model that most closely aligns customer usage with value. Zillow’s a good example of this model. Your first listing is free. Beyond that, there’s a $9.99 per listing per week fee. The tradeoff here is that while per unit pricing might be a great way to get people to dip a toe in and try a service, as their usage rises, it can get expensive.

2. Tiered pricing plus overage. The middle approach is a combination of tiered pricing plus overage fees. It works well in B2B or B2C settings where you expect high usage, the software industry’s annual true-ups are a good example of this. With overage pricing, companies charge a fee for usage over and above the tiered allowance. For example, PagerDuty, a leader in digital operations management, offers a variety of different plans, from Free to Business, and even an additional tier called Digital Operations, an “end-to-end digital operations solution tailored to you.” But if you need more global notifications than is allotted for in your plan, you can pay an additional charge per notification.

Another option is for companies to automatically move customers with higher than expected usage to the next tier. Companies can use this combo approach to set usage and pricing thresholds based on what actual usage looks like.

3. Multi-attribute pricing. In a multidimensional model, companies adjust pricing based on a number of factors. Octo Telematics (formerly Omoove), the European leader in technology solutions for the mobility market, provides a good example of this model, leveraging multi-attribute pricing to change pricing based on various attributes including type of car, day of week, time of day, miles driven, and garage location.

This isn’t an exhaustive list. The key is offering customers choice and flexibility. Whichever approach you use, the context should be what your customer journey looks, and what behavior you want to motivate.

Of course, it’s not all smooth sailing on the road to implementing usage-based pricing. There are some caveats:

1. Revenue recognition. One area of major impact is revenue recognition. We’ve seen finance teams act as gatekeepers preventing product teams from adopting usage-based pricing because of the impact. Because trying to manage usage-based pricing manually is so time-consuming, not to mention error-prone, it’s something you want to automate. And you need to have a plan to tackle it efficiently. But at the end of the day, scaling your processes to handle usage-based billing shouldn’t be used as a roadblock to moving forward.

2. Pricing agility. Konary spoke to the importance of having agility in pricing systems as well as building in flexibility. For instance, not 100% usage, but usage plus tiers—so customers can scale pricing up or down. Basically, it’s a balance between monetizing the usage and making sure they are interacting with the system or the service. You’re always trying to increase customer lifetime value. But without the right agile systems to support this pricing optimization, you’re likely to run into trouble.

3. Managing usage data. Everything from the complexity to the volume to the real-time component of collecting usage data—including alerting customers they’re close to hitting usage limits—drives up the difficulty in managing usage pricing. You need to have a process for data collection, handling, and aggregation to avoid data loss and the resulting revenue leakage. You also need to be able to capture all of that data in the right place at the right time, in real time. Hayden identified the data quality step as a critical point where revenue leakage occurs. This makes detecting and isolating data issues super-important for successfully moving to a usage-based pricing model.

Figuring out the right metric takes deep thought, and deep awareness of — and empathy for — the customer experience. And remember, just as the customer experience is a living thing that’s always changing, so too should your pricing. The “right” usage metric isn’t necessarily the right pricing for all time. Incorporating usage-based pricing into your revenue mix means you need to commit to staying flexible enough to pivot your processes and iterate pricing based on market feedback.