How to Price a Round on Anchor ARR

You are pricing a round and one customer's signed recurring revenue is doing the work. Blackbird Ventures and Abstract Ventures led a $100 million Series A for General Compute at an $800 million post-money valuation. The Series A was triggered by one customer's $100 million annual recurring revenue commitment. The decision is whether that anchor can carry the multiple, or whether it is only evidence. The test has four lines:

• Size of the signed recurring commitment.
• Concentration risk.
• Secured dependencies.
• Next-round bar.

You do not need a perfect answer on every line. You need a defensible answer on the line that sets the multiple.

Start with the signed commitment, not the pipeline

Measure the recurring dollars that are already contractually yours. Use annual recurring revenue from the anchor customer, with renewal terms, payment terms, and termination rights visible. Give the investor a number that can be underwritten without asking for a forecast. Price a short-term commitment as short-term. Show the weighted value and exit clause for a multi-year one. Keep the anchor separate from the other revenue. The other revenue can be strong. It cannot replace the anchor when the anchor is the reason for the mark.

Before the Series A, General Compute had a $60 million post-money valuation after its May seed round. The mistake is treating a large quote as ARR when the customer can walk after a short notice period. A quote is a sales event. A commitment is a financing input.

Treat concentration as the price of the multiple

Price the risk before the lead asks for it. Build a downside schedule that names the trigger, the cash impact, and the recovery path: what happens to revenue, gross margin, and cash when the anchor pauses, renegotiates, or fails. A material revenue hit means the multiple has to be lower, not higher. A lead will not pay for a story about a customer that has not signed. They will pay for a customer that has signed, plus a plan for what happens when that customer changes its mind.

CoreWeave shares hit a 12-month low when General Compute's Series A funding arrived. Elon Musk forecast that over 200 neoclouds would experience severe failures and consolidation. In that environment, an anchor needs a downside case. Concentration is acceptable when the investor can name the mitigation. That mitigation might be another customer in the same segment, a longer contract, a prepayment, or a hardware purchase that locks in capacity. Name the mitigation or the multiple is unsupported.

Lock the dependencies before you ask for the mark

Secure the inputs that make the anchor revenue real. Build a dependency map that shows who controls each input and what happens when the input slips: power, chips, data center space, financing, and delivery dates, each with a counterparty and a fallback. Keep the map short. A promise dependency makes the anchor ARR a promise. A map that needs a slide deck to explain why the revenue will arrive is not ready to carry the multiple.

General Compute made a $400 million chip purchase described as the first debt-financed inference-hardware transaction. The projection was that General Compute would deploy 75 megawatts in 2027 and that the company could access 200 megawatts of power. Those details show the commitment can be delivered. A founder can price on anchor ARR while the hardware is still unconfirmed, but the investor will price on the difference between the commitment and the delivery plan. Close that difference first.

Set the next-round bar before the term sheet

Define what the next round must show. The bar can be another anchor, a larger backlog, a lower concentration ratio, or a public market signal that the category is still expanding. A current round priced on a single customer needs the next round to prove the model is not a single customer. Write the bar in the data room, not just in the term sheet. State the dependency when the next round depends on another customer, power, or a public market signal.

CoreWeave reported $99.4 billion of revenue backlog in its first-quarter results, including a $21 billion commitment from Meta. The next-round bar is contracted value large enough to make the current multiple look conservative.

This article is general information, not investment or legal advice.