When I was a national sales manager at a telecommunications company, one government customer made a crazy number of conference calls during an election. We celebrated. The next month their spending dropped off a cliff and we panicked.
Same customer. Same product. The difference was the metric we charged on.
Your value metric is the unit your price counts: users, jobs, locations, messages, resolutions. Get it right and revenue grows every time your customer does. Get it wrong and your biggest customers become your least profitable ones.
How to use the Value Metric Finder
- Pick your type of business and tap the suggested metrics you want to test, or add your own. Five candidates is a good number.
- Score each one from 1 to 5 on the five tests. Be honest. A 3 means "fine, not great."
- Read the ranking and the advice on your top metric's weak spots.
The tool starts with the worked example from our value metric guide: job-management software for plumbers, electricians and HVAC contractors.
The five tests
In Chapter 4 of the book I wrote that a usage metric "needs to be value-based, scalable, flexible, feasible, and predictable." Here's what each one asks:
- Value-based: does more of this unit mean more value for the customer?
- Scalable: does it keep growing as the customer grows? I warned that a metric that "stalls once the customer reaches a certain level" means you give away more than you intend.
- Flexible: does it work for a tiny customer and a huge one?
- Feasible: can you measure it accurately and cheaply, without arguments?
- Predictable: can the customer estimate next month's bill?
Score = Value-based + Scalable + Flexible + Feasible + Predictable (out of 25)
A worked example
For the trades software, five candidates score like this: per job completed 22, per user 20, per active customer 20, % of invoice value 20 and per business location 17.
Per job wins because it's high on value, scales with the customer's business and works for a solo plumber and a 50-van contractor alike. Its weak spot is predictability (3/5), because job numbers swing with the seasons.
The fix isn't to drop the metric. It's to package it: sell plans by jobs a month (say up to 50, 200 and 600) so the bill only changes when a customer moves up a band. In the guide, that ladder grows revenue with the customer far better than $35 per technician.
Notice the % of invoice value. It scores perfectly on value, but customers can't predict it and you'd need their invoicing data to measure it. Great in theory, painful in practice.
How to read the result
Look at value-based and scalable first. They decide whether revenue grows without a new sale. Feasible and predictable can usually be fixed with packaging: bands, allowances, caps and alerts.
Two traps come up again and again. Hours: if you charge by the hour, getting faster costs you money. Seats: if your product helps customers do more with fewer people, per-seat revenue shrinks as you get better.
My rule is one primary metric everyone understands, plus at most one secondary meter for the thing that genuinely costs you money (usually AI usage).
What to do next
- Build your tiers around the winner in the Pricing Table Builder.
- See 40 real value metrics with 2026 price points in how to choose a value metric.
- Weighing users against usage? Read seat-based vs usage-based pricing.
- If your metric is consumption, go deeper with usage-based pricing and the four factors to get it right.
- Not sure which pricing model the metric belongs in? Try the Pricing Model Picker.
Frequently asked questions
What is a value metric?
A value metric is the unit your subscription price is based on, such as users, locations, contacts, transactions or resolutions. A good one rises as the customer gets more value, so your revenue grows without a new sale. It's different from your pricing model, which is the structure (flat, tiered, usage or hybrid) that counts the metric.
What are examples of good value metrics?
Slack charges per active user, Twilio per message and minute, Snowflake for compute credits, Fin per resolution and Xero per organization tiered by payroll headcount. For non-software businesses, think per pet for a vet plan, units serviced for a maintenance plan, or entities and transactions for an accounting firm instead of hours.
Can I have more than one value metric?
You can, but keep it simple. Use one primary metric that everyone understands, and add at most one secondary meter for something that genuinely drives your costs, such as AI usage on top of a seat. Every extra meter makes your price harder to explain and your customer's bill harder to predict.
How do I make a usage metric predictable?
Package it. Sell bands or tiers (for example, up to 200 jobs a month) instead of a raw per-unit price, include an allowance in each plan, and send alerts before customers hit their limit. Leadfeeder in the book does this with lead bands sold as monthly subscription tiers.
Should I change my value metric for existing customers?
Treat it like a price change. Model what every existing customer would pay under the new metric, find the ones whose bill jumps, and give clear notice with a reason. Many companies move new customers first and migrate existing ones at renewal, with a transition discount for anyone who'd pay much more.
Built on The Subscription Playbook
This tool turns frameworks from Robert Coorey's book into something you can use in minutes. Robert is the author of The Subscription Playbook and a Director at BPAG Accountants. More about Robert.