Applies when you are guessing at a price, priced on a gut feeling and feel stuck, don't know where the free line goes, or believe your users won't pay so you never charge. Produces a value metric, a free-to-paid packaging model, and a defensible number, from evidence rather than fear.
The three decisions, in order
Most founders skip to the third and wonder why nothing converts.
- Value metric — what you charge per. The most important call; get it wrong and no tier structure saves you.
- Packaging — what is free, paid, and enterprise, and what triggers the upgrade.
- The number — the actual price on each tier.
Price against the value the customer receives, never against your cost or your fear. Your infrastructure bill is a floor, not a strategy.
Are you ready to price?
Do not gate before you have proof people come back. If second-week retention is weak, a paywall just turns a leaky funnel into a smaller one; prove retention first, then monetize. The exception: if buyers are already asking to pay you, that is a green light at any stage. Stated willingness to pay is the strongest signal there is.
The value metric
Charge for the thing that grows as the customer gets more value. A good metric scales with their success (seats, active users, projects, events, calls, data, builds), stays predictable enough to forecast, and is legible in one sentence.
- Value from more people → per-seat, but watch for shared logins and bot accounts deflating it.
- Value from more usage → usage-based, with a floor and caps so the bill stays predictable.
Hybrid (a platform fee plus usage) is common and fine. Never put per-seat on a machine-value product, or pure usage on a collaboration product: you tax the exact behavior you want more of.
The free-to-paid line
For an open-source or self-serve tool, the free tier is acquisition, not charity. Give away individual value; charge for team, scale, and trust.
- Free: the core value for one developer or a tiny team, generous enough to live in their workflow. This is your distribution.
- Paid: what a company needs that a person does not — collaboration and seats, higher limits, single sign-on, audit logs, roles, compliance, support and SLAs.
- Enterprise: "contact us," wherever security review and procurement enter.
Trigger the upgrade at a moment of earned value ("you added a third teammate," "you crossed the usage threshold," "you need SSO"), never an arbitrary wall or a hidden feature the tool is useless without. Developers forgive a paywall on "my company needs this," and resent one on "the thing you advertised."
Finding the number
Never pick it alone in a room. In order of strength:
- Deflected willingness to pay — the "can we pay for this" messages you already got. Ask what it cost them to go without, and what a tier needs to look like to get approved.
- Value anchoring — price against what you replace and the time or money you save; capture a slice of the value delivered.
- Competitor anchoring — know the number already in the buyer's head, then earn a reason to be higher or lower. "Roughly the same but a bit cheaper" loses.
- The range question, asked in user interviews: "At what price is this too expensive to consider? So cheap you doubt it?" The gap is your range.
What good looks like
- The value metric fits in one sentence, grows with the customer's success, and stays predictable.
- The free-to-paid line falls on team, scale, and compliance, not on the core thing you advertised.
- The number came from evidence, not cost or gut.
- A healthy share of prospects say "that's a lot." If nobody ever pushes back, you are too cheap.
- The customer gets roughly ten times the price in value; below about three times, they churn.
Rules
- MUST price against the value the customer receives, not your cost.
- MUST prove people return before putting up a paywall, unless buyers are already asking to pay.
- NEVER ship more than three self-serve tiers, or a usage price with no cap.
- NEVER discount at the first "too expensive"; it trains buyers to push and signals you do not believe your own value.
1---2name: dev-tool-pricing3description: Use this skill when setting or fixing the price of a developer tool or technical product: choosing the value metric, drawing the free-to-paid line, structuring tiers, and finding the actual number. Produces a value metric, a packaging model, and a defensible price grounded in evidence. Triggers on "what should I charge", "how do I price this", "free vs paid", "pricing tiers", "value metric", "usage-based vs per-seat", "we're too cheap", "developers won't pay", "should I charge yet".4---56Applies when you are guessing at a price, priced on a gut feeling and feel stuck, don't know where the free line goes, or believe your users won't pay so you never charge. Produces a value metric, a free-to-paid packaging model, and a defensible number, from evidence rather than fear.78## The three decisions, in order910Most founders skip to the third and wonder why nothing converts.11121. **Value metric** — what you charge *per*. The most important call; get it wrong and no tier structure saves you.132. **Packaging** — what is free, paid, and enterprise, and what triggers the upgrade.143. **The number** — the actual price on each tier.1516Price against the value the customer receives, never against your cost or your fear. Your infrastructure bill is a floor, not a strategy.1718## Are you ready to price?1920Do not gate before you have proof people come back. If second-week retention is weak, a paywall just turns a leaky funnel into a smaller one; prove retention first, then monetize. The exception: if buyers are already asking to pay you, that is a green light at any stage. Stated willingness to pay is the strongest signal there is.2122## The value metric2324Charge for the thing that grows as the customer gets more value. A good metric scales with their success (seats, active users, projects, events, calls, data, builds), stays predictable enough to forecast, and is legible in one sentence.2526- Value from more **people** → per-seat, but watch for shared logins and bot accounts deflating it.27- Value from more **usage** → usage-based, with a floor and caps so the bill stays predictable.2829Hybrid (a platform fee plus usage) is common and fine. Never put per-seat on a machine-value product, or pure usage on a collaboration product: you tax the exact behavior you want more of.3031## The free-to-paid line3233For an open-source or self-serve tool, the free tier is acquisition, not charity. Give away individual value; charge for team, scale, and trust.3435- **Free:** the core value for one developer or a tiny team, generous enough to live in their workflow. This is your distribution.36- **Paid:** what a company needs that a person does not — collaboration and seats, higher limits, single sign-on, audit logs, roles, compliance, support and SLAs.37- **Enterprise:** "contact us," wherever security review and procurement enter.3839Trigger the upgrade at a moment of earned value ("you added a third teammate," "you crossed the usage threshold," "you need SSO"), never an arbitrary wall or a hidden feature the tool is useless without. Developers forgive a paywall on "my company needs this," and resent one on "the thing you advertised."4041## Finding the number4243Never pick it alone in a room. In order of strength:44451. **Deflected willingness to pay** — the "can we pay for this" messages you already got. Ask what it cost them to go without, and what a tier needs to look like to get approved.462. **Value anchoring** — price against what you replace and the time or money you save; capture a slice of the value delivered.473. **Competitor anchoring** — know the number already in the buyer's head, then earn a reason to be higher or lower. "Roughly the same but a bit cheaper" loses.484. **The range question**, asked in user interviews: "At what price is this too expensive to consider? So cheap you doubt it?" The gap is your range.4950## What good looks like5152- The value metric fits in one sentence, grows with the customer's success, and stays predictable.53- The free-to-paid line falls on team, scale, and compliance, not on the core thing you advertised.54- The number came from evidence, not cost or gut.55- A healthy share of prospects say "that's a lot." If nobody ever pushes back, you are too cheap.56- The customer gets roughly ten times the price in value; below about three times, they churn.5758## Rules5960- MUST price against the value the customer receives, not your cost.61- MUST prove people return before putting up a paywall, unless buyers are already asking to pay.62- NEVER ship more than three self-serve tiers, or a usage price with no cap.63- NEVER discount at the first "too expensive"; it trains buyers to push and signals you do not believe your own value.
swan-gtm