In the Transition to Usage-Based Pricing, Invest in Your Data for Greater Predictability

You know that feeling when you’re in an unfamiliar city and you get off at the wrong subway stop? I still remember years ago when I mistakenly got off at Central Station in Chicago. It was disorienting, to say the least: Turning around, trying to find one’s bearings, looking for something familiar to orient yourself with, then realizing that you’re not where you should be and should retrace your steps quickly!

For many people transitioning to usage-based pricing (UBP), this feeling is quite similar. With the shift to UBP, people can get lost with all the newness around them. We’re talking about:

• new unit economics
• new client relationship management needs
• new compensation packages
• and new corporate performance metrics to track

All of these changes, by the way, are based on the same data: what the client has consumed and what they might consume in the future. But how can you predict consumption rates and deliver usage patterns that give investors confidence? In order to improve and obtain greater predictability, companies must invest in their data.

This statement seems obvious enough but based on my interactions with many companies making the transition to UBP, there are common data points that are overlooked and need investment. At the crux is obtaining clean, actionable data from multiple sources to provide successful predictability. All the data scientists in the world can’t predict accurately if their data is incomplete or messy.

Accurately forecasting usage allows orgs to scale with efficiency

The goal of every organization is to be effective at using its available resources, aligning investments to what is needed at any given time. Erratic shifts cause an ineffective deployment of capital. A predictable consumption forecast will result in predictable revenue and a better alignment of investments in human and computing capacity to grow. A high variance between actual and forecasted usage means that resources are either overworked or potentially sitting idle.

The challenges that most organizations face in forecasting usage has a compounding effect:

1. Predicting when contracts will close. It’s challenging for your sales leader to consistently determine who will sign and when.
2. Knowing what those contracts will ultimately produce. Booked contract values are often sales’ best estimate of usage at a fully onboarded state.
3. Understanding when that usage starts. Delays in set up or implementation causes delays in usage.
4. Projecting how the ramp-up will look. How long it takes for the customer to get to optimal usage and the curve to get there tends to differ depending on the type or location of the client.
5. Tracking variability with ongoing usage. Once the client has reached optimal usage, what are the factors that can affect usage variability?

More data, more problems: finding the right data to predict usage

“Clean, actionable data” is easy to say but so difficult to obtain. Your solution provides your customers with a lot of value. At the same time, it captures a lot of data – most likely a ton of data!

In addition to monitoring and measuring the functions that your clients value, your system likely has a lot of audit logs, database transaction monitoring, and so on. At the core is the data that your organization needs to measure “usage” as defined by your pricing strategy.

Chances are, your system wasn’t originally built to present this data in an easy-to-consume way. Nor was it built to serve the needs of all those who require access in order to do their (new) jobs. These three areas of investment can provide an easier path to predictability:

Invest in having your system produce the data that you need

This kind of investment makes sense, but expect friction when implementing it. Companies are in a constant race for their customers’ attention and wallets. The focus always seems to be on the next cool feature or on flanking the competition’s capabilities.

As a result, companies deprioritize the changes required for optimizing the production environment to accurately capture key usage data. Investments in core infrastructure are often hard to implement. Cranking out critical data for driving usage-based pricing is hard, resource intensive, and frankly not very glamorous for engineers.

Invest in a universal and clean “usage data pipe”

Moving to UBP requires operational changes in every part of the business. Sales needs to focus on revenue as opposed to just contract bookings. Customer success needs to better understand what the customer is doing and present that data in meaningful dashboards to clients. Finance needs to have high confidence in the numbers and a scalable way to bill customers.

Each organization requires its own data feed and each will think that their needs are unique. It’s critical for everyone to agree on one view of the truth. This means that someone in the engineering team can understand each group’s needs and then translate these needs into one aggregated data set. Providing API access to this might be the first step.

One other important point about a universal and clean data pipe is its auditability. The data that we are describing will be used to accurately measure usage, which clearly impacts revenue. When constructing the data and its use, expect it to withstand the scrutiny of auditors, especially if the aspiration is to IPO one day.

Invest in data sources closest to the customer

Usage data from your core system is extremely important, but it’s not the only data source that will improve predictability. Forecast accuracy increases linearly with each additional clean data source you incorporate. Here are three approaches to consider:

1. Ask those closest to the customer for their input. Organizations consistently say that they have challenges in predicting the ups and downs of customers’ usage patterns. And yet, nearly all the ones that I have spoken to do not ask their teams serving those customers for input. The most common reasons cited are that it’s “too hard” or it “doesn’t scale.” This is curious, as it is just data after all. Find an easy way for sales, account managers, and customer success to log crucial customer usage data (like in your CRM) and let it flow.
2. Pre-sales engineers can help customers find the right usage tier. Some organizations employ pre-sales engineers to work with prospective customers to understand their previous transaction volumes and seasonality. This enables potential customers in selecting the appropriate usage and pricing tier. Analyzing prior usage is rich data that can help finance to model the customer’s first-year usage and revenue ramp. Organizations can harness this data with relative ease – but I have yet to see any do so.
3. Ditch spreadsheets as a means of gathering data. Some organizations have occasionally polled their field for their predictions. They typically do this for their large customers and end up aggregating the data in a spreadsheet. However, this approach is problematic for a few reasons:

• • it’s time intensive
  • it takes a lot of elapsed time
  • it causes “VLOOKUP hell” for finance
  • and it is almost always out of date by the time the analysis is complete.

There’s also another issue with this approach. The revenue forecast data is now in a highly portable format that creates a governance risk if it were to leak outside the organization.

Investing in data makes your transition to UBP easier

The transition to usage-based pricing is a journey. It’s one that is full of adventure and the odd surprise every now and then (like my stop at Chicago’s Central Station). Developing an investment roadmap for your data needs will be a boon in predicting revenue. It helps to smooth out any variability increases while providing valuable forecasting for your whole organization.