What do I charge, and on what metric?

You are pricing something your buyer has no budget line for yet. Every benchmark you would normally borrow, the analyst's range, the competitor's rate card, your last company's deal desk, describes a market that has already agreed with itself. Yours has not. That is not a handicap. It is the one moment you get to set the anchor instead of inheriting one, and most of what follows is about not wasting it.

Attach the price to a budget line that exists.

You are not asking for new budget. New budget for a problem nobody has named barely exists, and the people who hold security budgets are being told to consolidate vendors, not add them. You are attaching your price to a line that is already funded: endpoint, identity, telemetry, governance, or the services line where a consultant's statement of work is doing your product's job badly. The line you land on sets three things. Your ceiling is what that line spends today. Your signer is whoever owns the line. And your defense at renewal is that the owner defends their own line next year, while an invoice with no line to belong to has to defend itself.

The services line is often the honest first home. You are replacing a statement of work before you are replacing a product, and a buyer who cannot name your category can still name the consultant you make unnecessary. If no line exists at all, price so that a single director can sign alone. A price that needs a committee needs a category, and you do not have one yet.

Ask your champion two numbers before you write a proposal: what they can sign alone, and what dollar figure pulls in sourcing, legal, and a full security review. Every large buyer has a signature-authority matrix with those thresholds in it, often different for a new vendor than for an existing one, and that threshold, not your list price, decides how long you wait. Price the first order under the second number if you can live with the scope.

Pick the metric before the number.

You will get the number wrong and correct it later. The metric is the decision that sticks. It survives the repricing, the Series A, and the first competitor, and it is nearly impossible to change at renewal. Run three tests, in order. Does it scale with the buyer's exposure, the thing they are afraid of, rather than with your cost to serve? Can procurement count it without an argument, because a count that needs your dashboard to settle becomes a dispute you invoice every quarter? And does it avoid taxing the behavior your product needs? Per-gigabyte pricing taught a generation of security teams to drop telemetry, and the products priced that way are still resented for it.

The same argument is being run in public right now over products driven by AI agents, split between per-investigation, per-agent, per-gigabyte, and flat pricing, and no model has won. Run the three tests on the new meters too. A per-investigation price rewards suppressing alerts in exactly the way per-gigabyte pricing rewarded dropping logs.

Sell one package.

You publish one package until the market has a name. Good, better, and best tiers are the furniture of a settled category. They assume the buyer already knows what the product is and is only choosing how much of it to take. Before that, a menu invites the buyer to compare you against the adjacent category you are trying to leave. One product, one annual term, one line on the order form.

The word platform can wait too. A platform price is a bet on your roadmap that the buyer has not agreed to make. Charged flat, the fee gets mapped to the one module they actually use, and the rest reads as shelfware.

Never buy the first customer with a discount.

You will be told the first logo is worth almost any price for its reference value. The reference value is real. The price is more durable than the logo, and far harder to undo. The number you sign becomes the reference price your next buyer hears about, the floor your renewal negotiation starts from, and the figure an acquirer's diligence uses to model what you can actually charge. We have watched a first-customer discount follow a company all the way to its exit. Prefer a small, tightly scoped deal at a real price over a big-sounding logo at a fraction of list.

The pilot's own number and its ending belong to the design-partner agreement. Design partners pay, on the same principle: discounted from a printed list price, capped in number, with the end of the discount written into the contract. Their payment buys you a procurement path walked end to end while you are in the room, a renewal date that forces the value conversation, and a reference whose endorsement cost them something. If you run a free offer to open doors, print the list price on the paper anyway and discount to zero against it. An anchor you never printed is an anchor you never set.

Take the deal fast, and hold the price.

You take the money. At the first deals the thing worth optimizing is speed — get signed, get deployed, get the learning, and do it again — and almost nothing else you could win is worth the weeks it costs.

Speed and price are different arguments, and this is where founders collapse them. Moving fast means using the buyer's process, not arguing over a clause, and not waiting for a better quarter. It does not mean discounting to get there. A price cut buys a signature and costs you the anchor. A fast, clean process buys the same signature and costs you nothing. Optimize the calendar and hold the number.

Answer a discount request with scope.

A discount request from procurement arrives on nearly every first enterprise deal, as a percentage with no reason attached. The answer is scope, not price. Offer fewer assets, fewer identities, fewer sites, a shorter term, or a later start, and hold the unit price where it is. A discount with nothing removed teaches the buyer that the price was never real, and the next buyer will have heard. Give the rep no authority to discount in the first year, because the first request from every new seller is a discount, and it hides whether the deal has a demand problem or a qualification problem.

And ask what sourcing has to report. Sourcing is measured on savings against your first quote, so holding the price and cutting scope gives them nothing to write down. Give them something countable that costs you little, a multi-year price lock, extra training days, a delayed start, a named reference, and quote with room for one of those, never with room for a percentage.

Sell annual. Let renewals expand.

You sell an annual term, because a renewal date is the one moment the value conversation cannot be avoided, and you design the first contract to force a review of scope at that renewal. In our portfolio the largest account jumps came at renewal, not at the first sale, and the structure that worked turned expansion into a sequence of named projects the customer works through, rather than tiers or seats you push. Expansion built as tiers and seats is owned by the vendor and pushed uphill. Expansion built as the customer's own project list is pulled from their side of the table.

A multi-year contract is worth signing only when it is prepaid. Prepaid, it is runway that costs no equity. Not prepaid, it is a discount with a longer memory.

Cut the cost of buying before you cut the price.

You get paid after your buyer survives buying you: the security review of a security vendor, the questionnaire, legal, and the deployment window. That work is part of your price, and the buyer knows it even when you do not. Cut it before you cut the number. Answer the questionnaire before it is asked, put a trust center up before anyone requests a document, and use paper their counsel has seen before. A discount does not fix a hard purchase. It tells the buyer you are not sure either.

A private offer through a cloud marketplace belongs to the same job. It draws down budget the buyer's procurement already approved, which can turn a purchase from a new line into a drawdown of an existing one. It replaces the budget fight. It does not replace the security review.

Repricing in a slow quarter is busywork.

The temptation is to reopen pricing at every hard moment: at launch, when a new sales leader arrives, after a restructuring, and again as a stripped-down edition, and none of it changes the trajectory when the real problem is demand. We have watched companies reprice at every hard moment and never once ask whether anyone wanted the product. When deals are not closing, lowering the price or adding a cheaper tier is the standard first move, and it is the fastest way to make the demand question unanswerable for another two quarters. A miss is at least as often evidence of underpricing as of overpricing.

Raise price on scope, never on the metric.

Price increases get accepted when they come with more of what the buyer counts: more assets, more identities, more coverage. What they will not accept is a change to the metric itself, because the metric is wired into how they budget. Raise on scope every year and leave the metric alone.

The competition arrives twice.

Nobody stays alone at these prices, and the competition arrives in two waves. The funded field competes on price first, because price is the only thing a fast follower can change in a quarter. Then the platforms arrive with consolidation math and a bundle that puts a version of you inside a fee the buyer already pays. Your defense was chosen earlier than it looks. A metric wired into how the buyer already counts their own environment is hard to switch away from. A flat fee is a number anyone can undercut. And the budget line you attached yourself to is yours to defend at renewal. The metric is the part of the deal that was never supposed to move.