Startup Pricing: A Founder''s Guide to Setting, Testing, and Raising the Number That Decides Everything
Pricing is the single highest-leverage lever a founder controls. A 10% price increase, fully passed through, is often the difference between running out of cash in month 14 and closing a Series A. Most founders set their price once and never revisit it.
Cost-plus ("it costs us $12 to serve, so we charge $30") ignores what the customer is willing to pay. Competitor-matching ("Acme charges $99, so we charge $89") assumes the competitor priced correctly, which they almost never did.
Value-based pricing starts from a different question: what is the outcome the customer buys, and what is it worth to them? The standard for B2B SaaS is roughly 10–20% of the annual value delivered. Below 10% and you leave money on the table. Above 20% and the payback period gets uncomfortable for the buyer.
Per seat. Best when value scales with number of users (Slack, Notion). Predictable, easy to expand.
Per unit of consumption. Best when value scales with usage (Twilio, Snowflake). Aligns cost to value.
Per outcome. Best when the outcome is measurable and attributable. Highest willingness to pay, hardest to enforce.
Per platform tier. Best when different customer segments want different feature sets (HubSpot). Predictable at scale.
Pick one primary axis. Add a second only when the first stops explaining willingness to pay.
Starter. Priced to be adopted with no procurement. Usually under $5K/year. Feature-limited but not broken.
Growth. The tier most customers land on. This is where 60% of revenue should come from.
Enterprise. Custom quote. Includes SSO, SLAs, dedicated support. Anchored high so Growth looks reasonable.
Never remove the feature that made the product work. Gate on volume, seats, integrations, and support — not on the core loop.
1. Show the price. "Contact us" for the Growth tier kills conversion by 40%+. Reserve "Contact us" for Enterprise only. 2. Anchor with the highest tier on the right. 3. One-line value prop per tier, six differentiating features under it. Not 22.
Annual toggle should be 17–20% cheaper (two months free) and default on. This raises cash collected and reduces churn measurably.
Every discount is a permanent decision. If you give it once, it becomes the ceiling for the next negotiation.
Never discount without a concession. Longer contract, larger commit, published case study, co-marketing.
Never discount off list. Discount off an already-quoted price.
Approval matrix. AE up to 10%, VP Sales up to 20%, CEO above 20%.
End-of-quarter discounting to hit a number churns at 2x the rate of full-price customers.
1. Win/loss data on price. Of deals lost, how many named price as the primary reason? Under 20% means you are priced too low. 2. Discount rate. Average discount above 12% means list is too high or packaging is broken. 3. NRR by cohort. Above 120%, you have pricing power to raise the base. Under 100%, raising the base makes churn worse.
Output: one change. Not five. Ship it next quarter. Measure it the quarter after.
Once a year is normal. Twice is aggressive. Zero is a mistake.
Grandfather existing customers for 12 months. Turns a churn risk into a retention lever.
Announce 60 days ahead. In-flight deals close at the old price.
Give existing customers a reason. New features, expanded support, higher SLA. Never "our costs went up."
Typical result of a well-run 10% increase, twelve months out: net revenue up 6–8%, churn up 0.3–0.5% (statistical noise).
Pricing is a statement about how much value the product delivers. Set the price low, and the customer will believe the product is worth little. Set it at 10–20% of the value delivered, and both sides believe the same story.
Set the number. Package it in three tiers. Enforce the discount policy. Review every quarter. Raise every year.
