iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
The right evaluation separates customer concentration from supplier concentration, measures each with the correct denominator, and then tests whether losing the relationship would cause a severe near-term disruption. A large percentage is a warning signal, not a verdict: the decisive questions are whether the counterparty can leave or change terms, whether alternatives are qualified, and how long recovery would take.
For customers, examine revenue and accounts receivable by counterparty. For suppliers, examine purchases and accounts payable. Compare multiple periods, trace indirect relationships, and read contract and continuity disclosures before deciding whether the exposure is survivable.
What concentration risk means for an AI company
Concentration risk exists when too much of a company’s demand, cash collection, capacity, or operating capability depends on a small number of counterparties. In an AI business, that dependence can run in either direction:
Free tools Windows power users keep installed
One-click scans. No signup required.
- Customer concentration: a few customers, platforms, resellers, or public-sector buyers account for a large share of revenue or receivables.
- Supplier concentration: a few providers supply critical compute, data, model access, software, components, facilities, or other operating inputs.
These are different exposures. A company can have diversified customers but depend on one cloud platform, or sell mostly through several resellers whose ultimate demand comes from one model maker. Measure the two sides separately before combining them in an overall risk view.
#1 Best Overall
FASB’s concentration framework describes the vulnerability as exposure to a loss greater than the entity would have faced if it had mitigated the risk through diversification. The practical implication is that a percentage matters only when connected to a plausible loss event, its timing, and the company’s ability to respond.
Start with the company’s business model and reporting perimeter
Before calculating a percentage, identify what the company actually sells and which entities are inside the reporting boundary. An AI company might sell foundation models, applications, inference or hosting, accelerators, data-center equipment, data, or professional services. Each model has different bottlenecks and counterparties.
Map the commercial and operating chain
- List direct customers, distributors, resellers, cloud platforms, and managed-service partners.
- List providers of cloud capacity, colocation, data centers, accelerators, custom chips, networking, data, model access, software licenses, and other critical inputs that apply to the business.
- Consolidate entities under common control when the financial statements do so. A group of subsidiaries may be one economic counterparty even if invoices use several legal names.
- Mark each relationship as direct or indirect. The invoicing entity may be an intermediary rather than the ultimate source of demand or supply.
Do not assume that the contracting counterparty is the ultimate customer. An infrastructure seller may invoice a cloud service provider while end demand is generated by model developers, enterprises, or government buyers using that provider’s capacity.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Measure customer concentration with two separate indicators
Customer revenue and customer receivables answer different questions. Revenue shows where the period’s sales came from; accounts receivable shows how much unpaid cash is exposed at the reporting date.
| Indicator | Calculation | What it tells you |
|---|---|---|
| Customer revenue share | Customer revenue for the stated period ÷ total revenue for that period | Dependence of current-period sales on the customer |
| Customer receivables share | Receivable from the customer at period end ÷ total accounts receivable at period end | Collection exposure and working-capital dependence at a specific date |
Always record the numerator, denominator, currency, reporting period, and whether the figure is audited, estimated, or management-reported. A quarterly share, a six-month share, and a fiscal-year share are not interchangeable.
Use a multi-period trend, not one snapshot
Compare at least two comparable periods and, where available, quarterly data within the year. An annual decline can conceal a sharp recent increase, a customer loss late in the year, or a transition from one dominant customer to another. Track both the percentage and the absolute amount.
For every named customer, capture:
- Revenue amount and percentage for each period.
- Receivable amount and percentage at each period end.
- Whether the customer is direct, a reseller, or a platform intermediary.
- Contract duration, renewal date, termination rights, minimum commitments, pricing provisions, and exclusivity, if disclosed.
- Whether several named intermediaries may ultimately depend on the same buyer.
Interpret the 10% public-company disclosure trigger correctly
A presentation of FASB ASC 280-10-50-42 states that a public entity must disclose when one external customer provides 10% or more of revenue, along with the amount and the reporting segment or segments. The customer’s identity does not have to be disclosed. That is a financial-reporting disclosure trigger, not a universal safe-risk threshold for AI companies. A customer below 10% can still be operationally critical, and a customer above 10% may be replaceable if contracts and alternatives are favorable. Confirm the current authoritative FASB Codification text and effective date before using the rule for compliance or legal conclusions.
Rank #2
Measure supplier concentration by purchases and payables
Supplier analysis mirrors customer analysis but uses different measures. Purchases show how much of the period’s input spending came from a supplier; accounts payable shows unpaid obligations at the reporting date.
| Indicator | Calculation | What it tells you |
|---|---|---|
| Supplier purchase share | Purchases from the supplier for the stated period ÷ the filing’s stated purchase benchmark for that period | Dependence of input procurement on the supplier |
| Supplier payables share | Payable to the supplier at period end ÷ total accounts payable or the filing’s stated payable balance | Payment and liquidity exposure to that supplier at a specific date |
Read the denominator exactly as the filing defines it. Supplier purchase share is not automatically the same as a share of total company costs, cost of revenue, capacity, or physical supply. One SEC-filed concentration table reports a 96% supplier figure for the six months ended June 30, 2026 against a stated benchmark. It should not be restated as 96% of all company costs unless that denominator is established.
Identify the capability behind the supplier
A supplier of a noncritical office service is not equivalent to the sole provider of production inference. Classify each important supplier by the capability it controls:
- Cloud hosting, networking, storage, and managed inference.
- Data-center space, power, cooling, and physical operations.
- Accelerators, custom chips, memory, packaging, and other compute hardware.
- Training or inference data and data-rights licenses.
- Foundation-model access, APIs, orchestration software, and security tools.
- Specialized labor, maintenance, logistics, or other operating rights.
Record whether the supplier is a single source, one of several qualified sources, or merely the largest source. A large purchase share is more dangerous when the input is technically unique, capacity constrained, regulated, or difficult to qualify elsewhere.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Test resilience instead of stopping at the percentage
For every major customer and supplier, write a concrete disruption scenario. The test should answer whether the relationship can fail or change in the near term, what would be affected, and how quickly the company could recover.
Define the loss event
- Customer cancels, materially reduces usage, delays renewal, or demands lower pricing.
- Supplier raises prices, limits capacity, changes service terms, suffers an outage, or cuts access.
- Intermediary remains solvent but redirects demand to another provider or loses its own end customer.
- Contract expires during a period when replacement capacity or demand is scarce.
Estimate the immediate impact
Quantify the portion of revenue, cash collection, gross margin, delivery capacity, service availability, or customer commitments exposed during the first days, weeks, and quarter after the event. Separate accounting effects from operational effects: losing a customer may reduce revenue immediately but leave receivables collectible, while a supplier interruption may stop service before the income statement reflects the full loss.
Measure replaceability and recovery time
Ask whether qualified alternatives exist now, not merely whether other vendors operate in the market. Estimate the time needed to:
- Identify and contract with an alternative.
- Qualify hardware, software, data, or service performance.
- Move workloads, data, model weights, or integrations.
- Complete security, privacy, regulatory, and customer approvals.
- Reach the required scale and reliability.
Portability can be limited by proprietary APIs, data-egress charges, hardware-specific optimization, latency requirements, compliance boundaries, or model-performance differences. A nominally available substitute may therefore be unusable within the company’s recovery window.
Read the contract protections
Look for minimum purchases, take-or-pay commitments, renewal windows, termination rights, service-level remedies, exclusivity, price controls, capacity reservations, and change-of-control provisions. A customer contract that guarantees volume can reduce demand risk but increase exposure if the customer later fails to pay. A supplier contract with reserved capacity can improve continuity while imposing fixed costs.
Check operational preparedness
- Are workloads tested across more than one provider?
- Can data, model weights, and production images be exported in usable formats?
- Is spare compute or inventory available, and for how long?
- Are there independent routes to the same critical capability, or do several vendors rely on one upstream source?
- Has the company run a failover, migration, or customer-replacement exercise?
The ASC 275 concentration-disclosure approach focuses on whether a concentration exists at the reporting date, creates vulnerability to a severe near-term impact, and has a triggering event that is at least reasonably possible. Use that structure to keep the analysis tied to a specific scenario rather than an abstract score.
Trace indirect customers and shared ultimate counterparties
AI supply chains often hide concentration behind intermediaries. A company may sell to several cloud providers, neocloud builders, or resellers that all serve the same model maker or enterprise. Conversely, several apparent suppliers may depend on one upstream chip, foundry, data-center, or software source.
Build a chain for each material relationship:
- Identify the invoiced counterparty.
- Identify the known end user or upstream provider, if disclosed.
- Mark the relationship as confirmed, estimated, or unknown.
- Group counterparties that appear to share an ultimate buyer or source.
- Recalculate concentration under the grouped view and compare it with the legal-entity view.
NVIDIA’s fiscal 2026 filing illustrates why this matters: it says indirect customer revenue is estimated using purchase-order information, product specifications, internal sales data, and other sources, and describes a meaningful AI research and deployment company purchasing cloud services from its customers. Such figures should be treated as estimates and read alongside the company’s explanation of how they were derived.
Recommended Free Tools
What real filings can teach you
Issuer examples are useful for learning how to read disclosures, but they are not representative statistics for the AI sector.
| Disclosure example | Reported concentration | How to interpret it |
|---|---|---|
| SEC-filed annual report for fiscal 2025 | Three customers represented 29%, 11%, and 9% of revenue. | Revenue dependence was concentrated enough to require checking customer identity, contract terms, and trend. |
| Same fiscal 2025 report | Two customers represented 51% and 14% of accounts receivable. | Collection exposure did not necessarily match revenue concentration at the same date. |
| Same fiscal 2025 report | Two suppliers represented 55% and 36% of purchases. | Purchasing was highly concentrated; the relevant denominator was the report’s purchase measure. |
| Same fiscal 2025 report | Two vendors represented 63% of accounts payable and accrued expenses at year end. | Year-end payment exposure was distinct from purchase share. |
| SEC-filed concentration table for the six months ended June 30, 2026 | One customer represented 68% of revenue and another 10%; one supplier represented 96% of the filing’s stated benchmark. | These are period- and denominator-specific figures, not a statement about all costs or the AI industry. |
The fiscal 2025 issuer also said management believed alternative suppliers were available and that losing one would have a short-term effect. Preserve that statement as management’s assessment; it does not independently prove that replacement would be frictionless or immediate.
Compare AI companies on consistent axes
When comparing companies, use the same reporting period and denominator wherever possible. A useful comparison set includes:
- Largest-customer, top-three, and top-five revenue shares.
- Largest-customer and top-customer receivables shares.
- Largest-supplier and top-three supplier purchase shares.
- Supplier payables share and payment timing.
- Concentration trend, volatility, and turnover of named counterparties.
- Direct versus indirect exposure and shared ultimate counterparties.
- Contract duration, renewal and termination dates, minimum commitments, exclusivity, and pricing power.
- Availability of qualified substitutes, migration or qualification time, portability, and estimated recovery time.
- Scenario impact on revenue, gross margin, cash flow, capacity, and service continuity.
Do not collapse these dimensions into one unsupported industry score. A low percentage can be consequential when the input is irreplaceable. A high percentage can be manageable when alternatives are qualified, contracts are flexible, and migration has been tested.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallCrashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteA repeatable evaluation workflow
- Define the perimeter. Read the business description, segment notes, revenue disaggregation, major-customer note, supplier disclosures, and risk factors. List direct and indirect counterparties.
- Set the period and denominator. Write the exact dates and accounting measure before calculating any share. Do not mix annual revenue with quarter-end receivables or supplier purchases with total operating costs.
- Build the customer schedule. Record each disclosed customer’s revenue and receivable amount, percentage, direct or indirect status, and contract information.
- Build the supplier schedule. Record purchases, payables, critical capability, source status, and the filing’s benchmark for each material supplier.
- Compare periods. Calculate changes in both percentage points and absolute amounts. Investigate a falling total concentration if another customer or supplier is replacing the former leader.
- Map ultimate dependencies. Group intermediaries that appear to share an end buyer or upstream source, and label estimates explicitly.
- Run disruption scenarios. Model loss, price increase, capacity cut, outage, nonrenewal, or payment delay. Quantify near-term financial and operational effects.
- Test recovery. Document qualified alternatives, portability limits, contract protections, migration steps, and realistic recovery time.
- Assign a conclusion. State whether the exposure is low, watchlist, or acute only after considering share, severity, timing, and replaceability. Explain the evidence and the unresolved uncertainty.
How to read public-company disclosures and private-company gaps
Public-company filings may disclose a major-customer threshold, customer percentages, segment information, supplier concentration, and management’s view of alternative sources. The SEC Financial Reporting Manual also addresses customer trends and concentration disclosures. These notes often omit the customer’s name, detailed contract terms, exact supplier capacity, or the time required to qualify a replacement.
For private AI companies, the information gap is usually larger. Request the following directly when performing diligence:
- Revenue and receivables by customer for the last several comparable periods.
- Purchases and payables by critical supplier, with the denominator defined.
- Renewal, termination, minimum-volume, exclusivity, and pricing terms.
- Cloud, hardware, data, and model portability documentation.
- Capacity reservations, backup providers, inventory, and tested failover plans.
- Names of intermediaries and any known ultimate customers or upstream sources.
- Management’s estimated recovery time and the assumptions behind it.
Mark every estimate and disclosure gap. Absence of a named customer in a filing does not establish that the customer is immaterial, and an assertion that alternatives exist does not establish that they are available at the required scale, price, or performance.
Common analytical mistakes and their fixes
Treating 10% as a safety line
Fix: Use the 10% ASC 280 figure as a public-company disclosure trigger only. Evaluate operational severity and replaceability at every concentration level.
Using revenue alone
Fix: Add receivables to expose collection timing and supplier purchases and payables to expose input and payment dependence.
Best Value
Mixing denominators
Fix: Repeat the filing’s exact benchmark and period in every table and sentence. A supplier percentage against a stated benchmark is not automatically a percentage of total costs.
Assuming several intermediaries mean diversification
Fix: Trace ultimate buyers and upstream sources. Several legal counterparties may represent one economic dependency.
Counting theoretical alternatives
Fix: Require evidence of qualification, capacity, portability, approvals, and migration time. A vendor listed on a market map is not a ready substitute.
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Accepting management’s continuity statement without testing it
Fix: Preserve the statement, then examine contracts, prior migrations, failover tests, capacity reservations, and the time required to execute the plan.
How to monitor concentration after the initial review
Concentration changes faster than annual filings. Set a monitoring cadence that matches the exposure:
- Quarterly: update customer revenue, receivables, supplier purchases, payables, contract renewals, and concentration percentages.
- Monthly for critical compute or cloud: review utilization, reserved capacity, outages, price changes, and migration readiness.
- At every renewal or major product launch: reassess whether a new model, platform, or customer has changed the ultimate dependency.
- After a disruption: recalculate recovery time, substitute performance, and cash impact using actual results rather than the original plan.
Escalate when a customer share rises quickly, receivables age materially, a supplier becomes the sole qualified source, contract renewal enters a short window, or a backup path has not been tested at production scale.
Bottom-line decision rule
Call an AI company highly dependent on a customer or supplier only when three elements align: the measured share is material, a plausible near-term event would cause severe financial or operational harm, and replacement is unavailable or slow. Customer revenue and receivables show demand and collection exposure; supplier purchases and payables show input and payment exposure. The most credible assessment states the denominator and period, follows indirect counterparties, preserves management’s caveats, and explains exactly how the company would continue if the relationship changed.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsQuick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
