Pricing — the lever that moves everything

Cost-plus = compute your cost to deliver, mark up by X%, that's the price. Easy. Defensible. Caps your margin at whatever competitor pricing tolerates. The right answer for low-differentiation commodities.

Value-based = compute the customer's economic value from the outcome (revenue gained, cost saved, time recovered), price as a fraction of that value. Hard to do. Defensible at much higher margins. The right answer for software that solves a measurable business problem.

Most founders default to cost-plus because it's easier to calculate, then complain about thin margins. If your customer can articulate the dollar value of the outcome you produce, you should be priced as a fraction of that value — typically 10-30% of the documented annual benefit.

The transition from cost-plus to value-based happens when:

1. You have 5+ paying customers
2. You can document the value created (numbers, not adjectives)
3. New deals consistently say "this is worth more than you're charging"

You don't need a 50-person sample. You need 10 honest conversations with the right people. The Van Westendorp method, adapted for founders:

For each candidate customer, ask four questions:

1. "At what price would this be so expensive you wouldn't consider it?" (the ceiling)
2. "At what price would this start to feel expensive, but you'd still buy?" (the upper band)
3. "At what price would this feel like a bargain — good value?" (the lower band)
4. "At what price would this be so cheap you'd doubt the quality?" (the floor)

Plot the answers. The optimal price range is where the "too expensive" line crosses the "bargain" line. Most founders land their first price 20-40% below this range because they're talking to themselves.

What you DON'T ask: "Would you pay $X?" — they'll say no even when they would, just to negotiate. Always ask in their language ("expensive/cheap"), never yours ("would you pay").

Packaging tiers: Three works (small/medium/large). Two doesn't (no contrast). Four-plus confuses. The middle tier should be the one you actually want to sell — design pricing so 60-70% of new customers land there. The high tier exists to anchor; the low tier exists to qualify (and to let price-sensitive customers buy something instead of bouncing).

Price increases: Best practice is annual. Grandfather existing customers for 6-12 months to avoid churn shock. Always communicate the why ("we shipped 7 new features this year") not the what ("our prices are going up"). Increases of 10-20% rarely move churn measurably; increases of 50%+ do.

Anchoring: List your highest tier first. Annual price first, monthly equivalent second ($199/yr ($17/mo)). Show the saving on annual ("Save 2 months"). Stripe and Vercel do all three; they're not coincidences.

Free tier: only useful if it serves as marketing for the paid product. If most free users never convert, your "free tier" is actually "customer subsidy."