A value metric is the unit a subscription price is based on, such as users, locations, contacts, transactions or resolutions. A good one rises as the customer gets more value, so revenue grows without a new sale. Test each candidate on five factors: is it value-based, scalable, flexible, feasible to measure, and predictable for the customer? Pick one primary metric, and add a second only if you must.
Key takeaways
- Your value metric is what the bill grows with. It matters more than the headline price.
- Score candidates on the five tests: value-based, scalable, flexible, feasible and predictable.
- Per-seat pricing is under pressure because AI lets customers do more with fewer people.
- Pick one primary metric. Salesforce found buyers were 'afraid of all these metrics' when it ran several.
- Non-software businesses have value metrics too: pets, vehicles, rooms, entities, visits.
When I wrote the book, Procore, the construction management software company, had a bold line on its pricing page: "Explore unlimited plans for every builder and budget."
When I first read it, I had the same reaction you probably have. There must be a catch. How can a software company let a builder add as many people as they like and still charge a sensible fee?
The answer is that Procore doesn't charge for people at all. It charges an annual fee based on the volume of construction the customer does each year, plus the features they want.
A small builder pays less. A big builder pays more. And as the builder grows, so does Procore's fee, without a single sales call.
That unit, construction volume, is Procore's value metric. I wrote about it in Chapter Four of The Subscription Playbook. It's aged well. For 2025, Procore reported revenue of $1.32 billion, gross revenue retention of 95% and net revenue retention of 106%. And when AI started letting builders do more with fewer staff, Procore wasn't exposed, because it never charged for heads. (The full story is in the Procore case study.)
Let me show you how to choose yours.
What Is a Value Metric?
A value metric is the unit your subscription price is based on.
Per user. Per location. Per 1,000 contacts. Per transaction. Per resolved conversation. Per pet.
It's different from your pricing model. Your pricing model is the structure (flat, tiered, usage, hybrid). Your value metric is the thing that structure counts. Two companies can both use three tiers, but if one tiers by users and the other by transactions, they'll grow very differently.
A good value metric does one job: when your customer gets more value, they pay more, automatically. You don't need to upsell them. The bill just follows the value.
The Five Tests of a Good Value Metric
In the book I said your metric "must be a metric that grows consistently with your customers and helps you communicate the value of your service. It needs to be value-based, scalable, flexible, feasible, and predictable."
Here's what each of those means in practice.
| Test | The question to ask | Fails when… |
|---|---|---|
| Value-based | Does more of this unit mean more value for the customer? | You charge for logins, storage or other things the customer doesn't care about |
| Scalable | Does it keep growing as the customer grows? | It stalls at a ceiling, so big customers pay the same as mid-sized ones |
| Flexible | Does it work for a tiny customer and a huge one? | Small customers can't afford the entry point |
| Feasible | Can you measure it accurately and cheaply? | You need expensive tracking, or customers dispute the count |
| Predictable | Can the customer estimate next month's bill? | Bills swing wildly and finance teams panic |
That "stalls" point is the one people miss. I warned in the book that if your metric doesn't keep growing, "you may base your pricing on a metric that stalls once the customer reaches a certain level. That means you give more of your service than you might intend."
Worked Example: Scoring Five Candidate Metrics
Let's say you run job-management software for trades businesses: plumbers, electricians, HVAC contractors. (This is an illustrative example.) You're choosing between five value metrics.
Score each test from 1 (poor) to 5 (great):
| Candidate metric | Value-based | Scalable | Flexible | Feasible | Predictable | Total |
|---|---|---|---|---|---|---|
| Per user (technician) | 3 | 3 | 4 | 5 | 5 | 20 |
| Per job completed | 5 | 5 | 5 | 4 | 3 | 22 |
| Per active customer | 4 | 4 | 4 | 4 | 4 | 20 |
| % of invoice value | 5 | 5 | 5 | 3 | 2 | 20 |
| Per business location | 2 | 2 | 3 | 5 | 5 | 17 |
"Per job" wins, narrowly. It's the purest measure of value (every job is revenue for the tradie) and it grows with the business. Its weakness is predictability, because jobs swing with the seasons.
So here's how you'd fix that: sell job bands as tiers rather than charging per job. For example, up to 50 jobs a month, up to 200, up to 600. The customer knows their bill, and you still grow as they do. I show the maths for this on tiered pricing models.
Now compare what three customers would pay. (Prices are illustrative.)
| Customer | Technicians | Jobs a month | Per user at $35 | Per job tiers ($59 / $149 / $349) |
|---|---|---|---|---|
| Solo plumber | 1 | 40 | $35 | $59 |
| Small electrical firm | 4 | 180 | $140 | $149 |
| HVAC contractor | 8 | 550 | $280 | $349 |
Notice what happens if the HVAC contractor buys an AI scheduling tool and cuts from 8 users to 6. Under per-user pricing, your revenue drops $70 a month. Under job tiers, it doesn't move, because they're still doing 550 jobs.
That's the whole argument in one row.
Why Per-Seat Pricing Is Under Pressure
Per user was the default value metric for a generation of software. It's easy to understand and easy to bill.
But it has two problems.
First, it punishes adoption. Every extra person the customer adds costs them money, so they ration access. Procore's "unlimited users" removes that objection and makes it the default tool across the whole team.
Second, AI shrinks seats. WiseTech Global, the Australian logistics software company, saw this coming. In December 2025 it removed seat and user fees from CargoWise and moved to transaction-based "Value Packs." Its reasoning: "For SaaS businesses that monetize based on seats or users, AI will disrupt them... our revenue is not diluted by customers becoming more productive." By February 2026, about 95% of CargoWise customers were on the new model. I unpack the move in how WiseTech killed seat fees.
That doesn't make seats wrong. If the value really is per person (a design tool, a chat tool, a sales seat), seats can still be the right metric. Slack charges per active user, and Figma prices Full, Dev and Collab seats differently because each type of user gets different value. For the full debate, see seat-based vs usage-based pricing.
30 Value Metric Examples From 2026
Here's what real companies charge for, as listed on their pricing pages or in our research (checked October 2026). Prices are in US dollars unless noted, and some are third-party estimates where marked.
| Company | Value metric | Example price point |
|---|---|---|
| Procore | Annual construction volume | Custom, unlimited users |
| WiseTech CargoWise | Transactions | Value Packs, custom |
| Slack | Active users | Business+ $15 per user per month (annual) |
| Linear | Users | $10 to $16 per user per month |
| Figma | Seat type (Full, Dev, Collab) | Professional $16 / $12 / $3 |
| Xero (Australia) | Organization, tiered by payroll headcount | Ignite A$37 to Ultimate A$143+ a month |
| Netflix | Household, by streams and quality | Standard $19.99, Premium $26.99 |
| Spotify | Household size | Individual $12.99, Family $21.99 |
| Greencross Vets | Per pet, per year | Healthy Pets Plus A$660 a year |
| Panera Sip Club | Drinks per month (capped at 30) | $14.99 a month |
| Raylo | Devices rented | Phones from £6.95 a month |
| Twilio | Messages and minutes | US SMS from $0.0083 each |
| Deepgram | Audio minutes | Pre-recorded from $0.0043 a minute |
| AssemblyAI | Audio hours | Universal-2 $0.15 an hour |
| Exa | Searches | $4 to $12 per 1,000 requests |
| turbopuffer | Data stored, written and queried | Minimums from $16 a month |
| Pinecone | Storage plus read and write units | $50 a month minimum |
| Snowflake | Compute credits plus storage | $2 to $4 per credit; $23 per TB a month |
| GitHub Copilot | Dollar AI credits at model rates | Pro $10 a month |
| Lovable | Credits per message | Pro from $25 a month (100 credits) |
| ElevenLabs | Credits (about one per character of speech) | Starter $6 a month |
| Runway | Credits per second of video | Standard $15 a month |
| Midjourney | Fast GPU hours | Basic $10 a month |
| Atlassian | Users, plus pooled Rovo AI credits | 25 to 150 credits per user a month |
| Fin (formerly Intercom) | Outcomes (resolutions) | $0.99 per outcome |
| Zendesk | Automated resolutions beyond an allowance | $1.50 committed, $2.00 pay-as-you-go |
| HubSpot Customer Agent | Resolved conversations | $0.50 each |
| Stripe Billing | Billing volume | 0.7% pay as you go |
| Owner | Online orders | $249 a month + 5% per order, or $499 flat |
| Paraform | Successful hires | About 20% to 25% of first-year salary (third-party estimate) |
A few patterns jump out:
- Infrastructure prices on raw units (minutes, searches, gigabytes). Customers are technical and can predict use.
- AI apps price on credits, because the cost per task varies. The ones customers trust publish what a credit buys. More on credit-based pricing.
- Support AI prices on outcomes, because a resolution is easy to count and easy to compare with the cost of a human. See outcome-based pricing.
- Consumer subscriptions price on the household or the pet, because that's how families think.
Owner is worth a second look. It sells two plans, $249 a month plus 5% per order, or $499 flat, and tells restaurants they cost the same at $5,000 of monthly online sales. That's a value metric explained so simply that the pricing page does the selling.
Find your value metric in five minutes
Score your candidate metrics with the free Value Metric Finder and see which one grows with your customers.
Can You Have More Than One Value Metric?
You can. Lots of 2026 products do. HubSpot sells seats and editions plus a monthly credit allowance. Atlassian sells users plus pooled AI credits.
But be careful. Salesforce ended up running per-action credits, per-conversation fees, per-resolution fees, per-user licenses and a fixed enterprise agreement all at once. One Salesforce executive said European customers were "afraid of all these metrics." Fixed-price enterprise agreements grew from 16 to more than 120 in one quarter as buyers looked for something simpler.
My rule: one primary metric that everyone understands, plus at most one secondary meter for the thing that genuinely costs you money (usually AI usage). If you need a third, you probably need a simpler product. The hybrid pricing guide covers how to combine a base fee with usage.
Value Metrics for Businesses That Aren't Software
This isn't only a SaaS idea. Every subscription has a value metric, whether you chose it or not.
| Business | Common metric | Better metric to consider |
|---|---|---|
| Gym or studio | Per member | Per member, with class-pack tiers |
| Dental practice | Per patient | Per patient, with a family plan |
| Accounting practice | Hours | Entities, transactions processed or returns lodged |
| Cleaning business | Hours | Rooms, square meters or visits |
| Trades maintenance plan | Per call-out | Units serviced (each air-conditioner, each hot water system) |
| Vet clinic | Per visit | Per pet, by life stage |
| Agency | Hours | Deliverables, campaigns or channels managed |
Hours are the trap. If you charge by the hour, getting faster costs you money, and your income is capped at your capacity. That's why accounting firms are moving to fixed monthly fees built on scope, which I cover in value pricing for accountants.
The Four Most Common Value Metric Mistakes
- Choosing what's easy to count instead of what's valuable. Storage and logins are easy to meter. Customers rarely care about them.
- Choosing a metric that stalls. If every customer above a certain size pays the same, your biggest accounts are your least profitable.
- Choosing a metric customers can't predict. Raw usage with no caps creates bill shock. Bands and allowances fix it.
- Ignoring the metric's effect on behavior. Per-seat pricing makes customers share logins. Per-message pricing makes them send fewer messages. Make sure your metric encourages the behavior you want.
What to Do Next
Here's how to find yours this week.
- List five candidate metrics. Think about what grows when your customer succeeds.
- Score each on the five tests using the table above, or run them through the Value Metric FinderScore candidate value metrics and pick the one that grows with your customers.Open the free tool →.
- Model your real customers on the top two candidates. Who pays more? Who pays less? Does it feel fair?
- Package the winner into bands so customers can predict their bill.
- Check it against AI. If your customer's team shrank by a third, would your revenue follow?
If you're not sure whether to go seat, usage, credits or hybrid, try the Pricing Model PickerFlat, tiered, seat, usage, credits, outcome or hybrid? Find out.Open the free tool →. For the bigger picture, start at the subscription pricing strategy hub, and if you're leaning toward metering, read usage-based pricing next.
Read the whole book free
The Subscription Playbook as a free ebook and audiobook. Australian business owners can claim a free hardcover, shipped free.
Claim the free hardcover Get the ebookFrequently asked questions
What's the difference between a value metric and a usage metric?
A usage metric measures activity, such as API calls or logins. A value metric measures something the customer cares about, such as messages delivered, jobs completed or issues resolved. Sometimes they're the same thing. The test is whether your customer would happily tell their boss, 'we used more of this, and it was worth it.' If they'd be embarrassed or annoyed, it's a usage metric, not a value metric.
What is the most common value metric in SaaS?
Per user (per seat) has been the most common for years, because it's simple and every buyer understands it. It's losing ground. Among the 44 companies that crossed about $100 million ARR in 2026, only 4 rely mainly on seats, while 18 of the 33 with visible pricing charge mainly on consumption. Seats still work well where value really is per person, such as design and chat tools.
Can I change my value metric later?
Yes, but it's one of the hardest pricing changes to make, because some customers will pay more and some less. WiseTech managed it by moving customers to transaction-based Value Packs that bundled new AI features, and about 95% of CargoWise customers had switched within a few months. Give long notice, show customers their own new bill, and protect existing contracts where you can.
Should I charge per user for AI features?
Be careful. AI features can cost you money every time they run, and a heavy user can cost many times more than a light one. Many 2026 products charge a seat price that includes an AI allowance, then sell credits or usage beyond it. Atlassian, for example, includes 25, 70 or 150 Rovo credits per user per month depending on the plan, pooled across the site.
What is a good value metric for a service business?
Look for the unit that drives both your workload and your client's value. For an accounting practice that might be entities, transactions processed or returns lodged rather than hours. For a cleaner, rooms or visits. For a vet, the pet. Avoid hours where you can, because hours reward slowness and cap your income at your capacity.
Get both books free
The Subscription Playbook (ebook + audiobook) and Feed a Starving Crowd, instantly. In Australia? Claim the free hardcover.
Robert Coorey, MBA
Robert is the author of The Subscription Playbook and Feed a Starving Crowd and a Director at BPAG Accountants in Sydney. More about Robert.
Sources
- Procore: fourth quarter and full year 2025 results (12 Feb 2026)
- WiseTech Global: 1H26 results briefing scripts (Feb 2026)
- Pricing Innovation: Agentforce's pricing journey (6 Oct 2026)
- Intercom (Fin) pricing, checked 27 Oct 2026
- Atlassian Support: Rovo usage limits
- Owner pricing, checked Oct 2026
Facts and prices checked 27 October 2026. Spotted something out of date? Tell us.