Pricing before you have pricing power | The PMF Playbooks

Early pricing feels like a guess because it is one. That is fine. The mistake is not guessing wrong; it is not guessing at all. Founders who delay pricing "until the product is ready" delay the only feedback that matters, which is whether a stranger will hand over money. This playbook is how to make a good first guess, learn from it quickly, and raise the price on purpose rather than by accident. The examples come from founders on The Product Market Fit Show.

Charge from the first customer

Free is not a price. It is the absence of a signal.

Free users do not tell you the truth. They stay polite, use the product lightly, and leave without telling you why. Paying customers complain, and complaints are the roadmap.
Free pilots do not convert. The person who agreed to a free pilot did not have to get anyone's approval. When the pilot ends and you ask for money, the approval process starts from zero, and most of the time it never finishes.
Free trains the market. If your first ten customers pay nothing, the eleventh will expect the same, and each of the first ten will resist paying later because they already have it.
Free hides the segment. Everyone takes free. Only the people with the real problem pay. Price is how you find them.

Anada Lakra launched BoldVoice, an accent-coaching app, in the first weeks of Y Combinator and charged $10 a month from day one with a one-week trial. Her reason was not revenue. Charging was the fastest way to find out who the customer was. The company is now at $10M ARR.

Charging for the product is a way to already figure out who that customer is.
Anada Lakra, BoldVoice, The Product Market Fit Show

Justin Adams, who sold his first company for eight figures and grew Aiwyn to $3M in a year, has never run a free pilot. His argument is psychological: people value what they pay for differently from what they get free, even when the price is one dollar. Once someone has paid, they feel bought in and they follow through. Justin Adams, Aiwyn

Soham Mazumdar says Rubrik, now a $16B public company, always ran paid pilots, because they wanted the customer to have some skin in the game. Saurav Chopra at Perkbox used paid pilots to fund development while proving value, on the way to $36M ARR and a $170M exit. Soham Mazumdar, Rubrik · Saurav Chopra, Perkbox

Pricing is part of the product

Founders think of pricing as a number they attach to a finished product. Harish Abbott thinks of it as part of the fit. He built Deliverr, e-commerce fulfilment that Shopify bought for $2.1B, and says the turning point was not finding product-market fit but finding what he calls product-price-market fit. Deliverr moved to a flat, pay-as-you-go price that included shipping, so a merchant knew exactly what an order would cost wherever it went. That change, plus a delivery badge, took the company from ten customers to $100M ARR in about two years.

Your product could be fitting, but the way you price it may not create the fit and so you have to deeply think about, am I pricing it to really enhance my product market fit?
Harish Abbott, Deliverr, How Deliverr got to $100M ARR through pricing changes

Mateo Marietti at CookUnity shut down $2M of on-demand revenue to restart at zero on weekly subscription meal plans, because a small group of customers had started ordering several meals at once. The subscription model had better unit economics. CookUnity now does $750M ARR. Merrill Lutsky at Graphite doubled revenue overnight with a single pricing change. CookUnity's pivot · Graphite: $1M ARR in 10 months

The lesson: treat the pricing model as a product decision you can test, not a label you apply at the end.

Find the value before you find the price

You cannot price what you have not measured. Before you set a number, answer, in the customer's own units:

What does the problem cost them today? Hours, salaries, late fees, lost deals, churned customers, fines.
What do they pay for the current workaround? A tool, a contractor, a share of someone's time.
What would they gain if the problem disappeared? New capacity, faster cycles, fewer mistakes.

Get these numbers from discovery calls, not from a spreadsheet you built alone. Then write one sentence: "For a [customer], we replace [workaround] that costs about [amount] and we [outcome]."

Dean Sysman built Axonius, cybersecurity asset management, to $100M ARR in four and a half years. Axonius prices against quantified outcomes, not a feature list: a commissioned ROI study, and numbers customers publish themselves, such as Western Union's 50 percent reduction in workload. When there is no reference price in your category, you have to manufacture the anchor from the outcome. How Axonius hit $100M ARR in 4.5 years

A useful rule of thumb: aim to capture roughly 10 to 20 percent of the value you create, and make sure the customer sees the other 80 to 90 percent clearly. If you cannot state the value in their units, the price will always feel too high.

Value to price

A founder selling AI triage for support tickets learned on discovery calls that a 12-person support team spent about a third of its time on routing and first responses. At an average loaded cost of $65,000 per agent, that was roughly $260,000 a year. They priced at $2,500 a month, a little over 10 percent of the value, and the buyer's reaction was relief, not resistance. The first draft of their pricing, based on "what similar tools charge", had been $400 a month.

Anchor high

Discounting from a high anchor preserves your price. Discounting from a low one erodes it. Andrew Rubin at Illumio sold seven-figure contracts to banks from day one, before the company had revenue to speak of. Amar Varma won an eight-figure deal for Autonomic by asking for so much money that only the CEO could sign it. The size of the ask made him part of the transformation rather than a vendor the company could ignore, and it led to Ford acquiring the business. Amar Varma, five-time founder

There is one deliberate exception. Amplitude launched against Mixpanel, which was first to market and better funded, with a product that was roughly ten times cheaper. Amplitude is now about twice Mixpanel's size and a $1.5B public company. Undercutting works when the incumbent's price has drifted far above its value. It is a strategy you choose, not a default you fall into. Jeffrey Wang, Amplitude

Pick a metric the customer already counts

The pricing metric is the unit you charge for. Get it right and pricing scales with value automatically. Get it wrong and every growing customer becomes a renegotiation.

A good metric is:

Already counted by the customer. Seats, tickets, invoices, locations, documents. If they have to learn a new unit to understand their bill, they will not trust it.
Correlated with the value you create. If you save time per invoice, price per invoice or per volume band, not per user.
Predictable. Customers hate bills they cannot forecast. A usage metric needs bands or a cap. Deliverr's all-in price won because merchants could predict it.
Matched to time to value. Wes Bush, who wrote the book on product-led growth, points out that if it takes a customer a quarter to launch your product inside theirs, billing as if value arrives on day one will not survive. Wes Bush on product-led growth
Hard to game. Per-seat pricing invites shared logins. Per-outcome pricing invites arguments about attribution.

For most early B2B AI products, a flat monthly fee per company in two or three volume bands is the right start. It is easy to approve, easy to explain and easy to change.

Set the number

With the value known and the metric chosen, pick the number with three checks.

The value check. Is the price clearly less than 20 percent of the measurable value, so the customer can defend it in one sentence?

The approval check. Who can sign at this price without a committee? In most companies a manager can approve a few hundred dollars a month, a director a few thousand, and anything above that goes to procurement. Price your pilot just under the threshold of the person you are talking to, unless you are deliberately pricing for the CEO's attention, as Varma did.

The seriousness check. Is it high enough that the customer will use the product? A $50 a month tool gets forgotten. A $1,500 a month tool gets a champion.

Then present it as a range you have already decided to be in, not a question. "It's $1,500 a month for the pilot, then $2,500 at the full tier." The Van Westendorp questions, asking customers what price would be too cheap to be credible, a bargain, expensive but worth it, and too expensive, are a fast way to find the range in a handful of calls.

Or test your way there. Mark Ang at GoBolt, the logistics company, ran successive price tests until demand told him where the ceiling was, and landed at $90 an hour.

Package in three tiers, and mean it

Three tiers work because buyers compare, and three gives them something to compare against.

Tier	Who it is for	What it includes	Purpose
Starter	The smallest real customer	The core outcome, limited volume	Lets a manager say yes alone
Team	Most customers	Full volume, integrations, support	The one you want them to buy
Business	Larger or regulated buyers	Security, SSO, audit, SLAs, onboarding	Pays for the things enterprises demand

Make the middle tier the obvious choice. Put the things every real customer needs in it, and keep the starter tier honest but limited. Do not build a tier you cannot explain in one line.

The pilot that leads to full price

A paid pilot is the right first purchase for almost every B2B product before product-market fit. Structure it so it converts:

Fixed term, 60 to 90 days. Long enough to see value, short enough to keep urgency.
A price the champion can approve alone. Usually 40 to 70 percent of the monthly full price, as a one-off pilot fee.
A success criterion agreed in writing. Measurable, and something you are confident you can hit.
The full price visible in the same document. Nobody should be surprised at the end.
A review at day 45. Show the numbers, fix anything that is off, and pre-agree the conversion before the pilot ends.

Accrual, AI for accounting firms, converted 100 percent of its pilots into production contracts with firms like H&R Block. The founders credit transparency during the pilot, not discounts. How Accrual found PMF

Take cash over headline ARR

Before product-market fit, cash matters more than a tidy ARR number. In the show's founder-led sales dataset, more than 60 percent of founders who closed their first $500,000 of ARR did it with annual prepayment, even at a 20 to 30 percent discount. SolidRoad's founder flew to customers to convert month-to-month plans into annual and multi-year contracts, and the company's net revenue retention reached 186 percent. Anada Lakra says defaulting BoldVoice to annual plans gave instant payback on acquisition and changed the unit economics. Founder-led sales: how 10 founders closed their first deals

Raising the price

You will underprice at first. Everyone does. Josh Reeves at Gusto surveyed customers on why they loved the product. Time saved, ease of use and simplicity came top. Price came about sixth. Some customers told him directly.

You should raise your prices. You're too cheap.
A Gusto customer, as recounted by Josh Reeves, How Gusto got to $1B in revenue

Reeves took that as a sign of fit. Knak, a no-code email builder for marketing teams, went from $100 a month to $5,000 a year, which reframed it from a utility to a business tool.

Raise on purpose, on a schedule, and tell people why.

Raise for new customers first. Every ten new customers, add 20 to 30 percent to the list price for the next ten and watch the close rate. If it does not fall, raise again.
Grandfather early customers for a fixed period, six or twelve months, then move them to the new price with notice. They took a risk on you. Say so when you write to them.
Tie the raise to something real. A new capability, a higher volume band, a security certification.
Never raise silently. A price change found on an invoice ends the relationship. A price change explained in a call, with a month's notice and an honest reason, usually does not.

Expect the cost to show up in growth, not churn. Anurag Goel raised Render's prices in January 2023. It was unpopular even inside the company. Churn was minimal. Growth slowed. He judged that not raising meant not running a sustainable business, and Render went on to $10M ARR with no marketing spend. Render: $10M ARR with $0 marketing

The signal to raise is that nobody is saying no. If every prospect accepts the price without hesitation, you are cheap. A healthy price gets a pause and then a yes from most, and a no from some.

Discounts: rules before you need them

Discounting is where early pricing falls apart. Decide the rules now and write them down.

Discount for something in return, never for nothing. Annual prepayment, a case study, a reference call, a logo on the site, a longer term.
Cap it. Twenty to thirty percent for annual prepay is the show's norm. Fifty is a different price.
Never discount the pilot below the level that keeps the customer serious.
Say no to the "one-time" exception. There are no one-time exceptions in pricing, only the new price for that customer forever.

Signs your pricing is wrong

Everyone says yes instantly. Too low.
Every deal ends in a negotiation. Metric or packaging is unclear.
Customers use far more than they pay for, and you resent it. The metric is not tracking value.
Customers churn at renewal despite using the product. The price was fine but the value was never shown. Fix the reporting before the price.
Sales cycles are long and involve procurement at small companies. Your price crossed an approval threshold you did not know about. Move the pilot under it.

Do this week

6 actions

01Write the value sentence for your main customer in their units, using numbers from real calls.
02Pick a pricing metric the customer already counts. Default to a flat fee per company in three volume bands.
03Set the pilot price under the champion's approval threshold, with the full price in the same document. Never make it free.
04Offer annual prepayment at a fixed discount and take the cash.
05Add the four Van Westendorp questions to the end of your discovery script and log the answers.
06Write your discount rules on one page and stick to them.

Pricing is the loudest question you can ask a market, and the only one it cannot dodge. Ask it early, listen to the answer, and keep asking as you grow.