Summary

  • Credo reported US$479.003m of fiscal-Q1 2027 revenue. Contracting Customer A and Customer B represented 43% and 28% of revenue respectively, while the same two anonymous counterparties represented 57% and 28% of accounts receivable.
  • The company also reports a distinct end-customer-profile view because some end customers have contract-manufacturing partners place orders. Four anonymous end-customer profiles accounted for 28%, 33%, 13% and 10% of quarterly revenue.
  • The filing does not provide a crosswalk between those tables. Its US$4.2m RPO, US$313.051m inventory, US$62.915m of capacity-reserve deposits and US$212.717m of remaining-fiscal-2027 manufacturing commitments are separate measures, not a way to reconstruct one.

A customer can be the buyer on paper and not the demand centre

There are two ways to read a customer concentration footnote. The quick way is to treat every percentage as a share of the same underlying buyer population. The more useful way is to ask what relationship each percentage measures.

Credo makes that distinction unusually explicit in its Form 10-Q for the quarter ended 1 August 2026. In the financial-statement concentration table, Customer A accounted for 43% of revenue and Customer B for 28%. Together, that is 71% of the US$479.003m reported revenue. The same table places 57% and 28% of accounts receivable with A and B: 85% of the reported US$288.798m receivable balance in aggregate percentage terms.

That is a meaningful contractual map. It shows which anonymous parties occupy the sales and receivable relationship the financial statements recognise. It does not tell readers that these are the only economically important demand centres, or that each is the owner of the data centre deployment eventually using Credo connectivity.

The company then supplies a second table. It says that it reports financial-statement revenue by the contracting parties that place purchase orders or sign revenue contracts, but that some end customers have contract-manufacturing partners place the orders. In those cases the contract manufacturer, rather than the end customer, is reported as the customer for financial reporting. To show a different form of concentration, Credo gives an end-customer-profile view: Customer C represented 28% of revenue, D 33%, E 13% and F 10%. The four profiles together equal 84%.

The point is not that 71% is false and 84% is true. Both are accurate sums within their own tables. They are measurements of different relationships. One follows the party that contracts, orders and can become the receivable counterparty. The other follows the end-customer profile that management believes better describes demand concentration. The filing gives neither names nor a mapping from A and B to C through F. It does not state whether an individual end customer appears through one intermediary, several intermediaries, or a mixture of direct and indirect ordering.

That missing bridge is commercially important. A reader who says “Customer A is Customer D” might be right, wrong, or only partly right; the public record cannot decide. A credit manager focused on collecting the US$288.798m receivable balance needs the contractual map. A procurement or capacity planner asking where demand is concentrated needs the end-customer-profile map. Neither task is completed by relabelling the other table.

The receivable is a different exposure from the revenue

The second distinction is temporal. Revenue says product was recognised in the quarter. Accounts receivable say a payment obligation remained at the balance-sheet date. Credo reported US$54.503m of cash use from the quarterly increase in receivables, primarily due to higher sales. That movement does not establish a deterioration in collections; higher sales can naturally produce more receivables. Nor does it say which customer or end-customer profile generated the change.

Still, the 57%/28% receivable table tells a reader something that the 43%/28% revenue table cannot. The largest contractual counterparties were not merely an historical sales mix. They also represented most of the aggregate receivable exposure at that date. That is why it would be a mistake to replace the contracting table with the end-customer-profile table in a credit discussion. An end customer can shape demand without being the entity that owes the invoice.

The reverse mistake is equally tempting. Calling the 85% receivable concentration “hyperscaler dependence” would invent identities, payment terms and a legal relationship that the 10-Q withholds. Credo says it serves hyperscalers, neoclouds, OEMs, ODMs, optical-module manufacturers, enterprise and HPC markets, but those market labels are not a key for decoding Customers A through F.

Inventory and reserved capacity are a third map

The working-capital disclosures show why identity and timing cannot be solved by one concentration statistic. Inventory was US$313.051m at 1 August, up through a US$61.547m working-capital use during the quarter. Management attributes that increase to support for unfulfilled backlog and related new-product ramps. The filing does not give the backlog amount, a customer-level allocation, a product-level inventory split or a promise that a particular build converts to revenue.

Capacity is reported on its own contractual terms. Other current assets include US$62.915m of refundable deposits for capacity-reserve arrangements, up from US$39.932m at the preceding year-end. The agreements with certain assembly subcontractors run for two to six years; deposits are refundable through future purchases upon meeting volume requirements. That description is neither an expense nor a completed recovery. It is a conditional right tied to future purchasing.

The company also reports US$212.717m of purchase commitments to manufacturing vendors and foundry partners for the remainder of fiscal 2027, alongside US$9.495m of technology-licence fees. These are supplier-side obligations. They should not be netted against an end-customer percentage, treated as a forecast of recognised revenue or assigned to a named buyer. They reveal that the balance between external demand and committed supply must be managed across several relationships whose names and terms are mostly not public.

Why the US$4.2m RPO does not complete the picture

The RPO number provides a useful negative check. Credo reported approximately US$4.2m of contracted but unsatisfied performance obligations, expected to be recognised over the following 12 months. That metric is small beside the quarter's US$479.003m revenue, but it is not evidence that demand is small, that there is no backlog, or that the supplier commitments lack support.

The filing defines RPO narrowly: it is transaction price allocated to unsatisfied or partly unsatisfied performance obligations, including unearned revenue and future invoicing. A product business with purchase-order and manufacturing pathways can have material recognised shipments, receivables, inventory and supply commitments without a RPO number designed to be a total order book. The public filing itself supplies the right discipline: keep the contract accounting measure inside its definition.

For the same reason, the quarter's SEC-furnished earnings exhibit is evidence of performance, not a decoder ring. It reports 114.7% year-on-year revenue growth, US$129.425m GAAP net income and US$764.3m of cash plus short-term investments. Those facts are relevant context, but they do not identify the customers, prove capacity recovery or convert the disclosed concentration tables into a single demand forecast.

A better reader's map

The defensible reading therefore has four panels.

First, contractual revenue concentration: A and B together account for 71% of reported quarterly revenue. Second, contractual credit concentration: those same labels carry 85% of receivables in the published percentage table. Third, end-demand concentration: C–F are four profiles totalling 84% of revenue under a distinct management view. Fourth, supply and timing: inventory, conditional capacity deposits, purchase commitments and RPO describe different operational claims.

That arrangement does not make the business less concentrated or more concentrated. It makes the question precise. Concentration is not a single physical object. It is a set of relationships—who authorises an order, who receives the product, who owes the invoice, who holds capacity, and when each commitment becomes cash or revenue. The missing crosswalk is not an inconvenience to be filled with inference. It is the boundary that preserves the meaning of the disclosures.

Sources