Customer Concentration Due Diligence for Competitor Bids

Executive Summary

Customer concentration due diligence for a competing bidder needs separate decisions about verification and disclosure. The seller should define the buyer's questions, have approved professionals reconcile restricted evidence, and release only a separately reviewed decision report. FTC guidance recommends minimizing information, evaluating what documents reveal together, and counsel review of clean-team outputs. These are context-dependent safeguards, not an automatic permission to disclose.

The recommended customer-evidence pack contains a defined period, a ledger-to-financial-statements bridge, a restricted identity key linking accounts to documented economic groups, and an exception schedule. PCAOB audit-evidence principles emphasize accuracy, completeness and sufficient precision; they are useful analytical disciplines here, without turning a transaction review into an audit. ICAEW similarly describes reconciliation of management information in financial due diligence. Buyer-oriented guidance also recommends grouping commonly owned affiliates. ^{[1] [2]}

In the Hypothetical Example, a distributor has invented annual revenue of $24 million and gross profit of $6 million. Separately billed affiliates contribute $2.4 million and $1.2 million of revenue, with $0.8 million and $0.4 million of gross profit. Together, their booked attribution is 15% of revenue and 20% of gross profit, calculated from the invented inputs. An additional $120,000 of unallocated credits is already included in the company totals. Its unresolved customer allocation prevents a definitive group-margin conclusion and may change the group's net revenue share. Evaluating profit concentration alongside revenue is a recognized diligence approach; the specific figures here are original illustrative calculations, not external findings or safe thresholds. ^[3]

Restricted raw evidence and bidder findings are different outputs. A transaction-specific SEC-filed agreement illustrates separate storage and approval of reports before wider distribution; it is not a universal protocol. A proposed recipient's actual pricing and strategy responsibilities matter, while coded records can remain identifiable. The practical deliverables are a request matrix, a reconciled restricted analysis, a minimized report and a release register recording versions, recipients, approvals and limitations. Customer calls, consent requests and announcements require their own decisions. Transaction and antitrust counsel should design the actual controls, and accountants should establish the financial basis. ^{[4] [5] [6]}

Introduction and Background

Customer concentration due diligence asks how much a business depends on particular customer relationships and whether the associated revenue and profit can endure through a transaction. For a seller facing a competing bidder, the question also concerns disclosure: what evidence answers the bidder's investment questions without releasing unnecessary customer intelligence? Buyer-oriented diligence already covers revenue quality, retention, contracts and transfer risk. The seller needs a controlled way to prove those attributes. ^[2]

This report addresses US owners and chief financial officers (CFOs) of business-to-business (B2B) manufacturers, distributors and service companies. It is written for Laurent Capital's owner and CFO audience, with a narrow decision in view: approving a specific customer-evidence pack with transaction and antitrust counsel. It is an educational process reference, not legal, investment, tax or accounting advice. The proposed worksheets are analytical tools for those advisers to adapt, rather than a legal protocol.

An evidence pack here means a defined source file, reconciliation, grouping method, exception record and proposed release. Its purpose differs from creating a persuasive presentation. The CFA Institute distinguishes financial reporting quality, concerning disclosed information, from the quality of the underlying earnings. A polished customer chart cannot establish the durability of the economics it depicts. ^[7]

The seller's competing objectives are credible verification and controlled exposure if the bid fails. A non-disclosure agreement (NDA) belongs in that process, but access and permitted use need separate attention. FTC guidance calls for minimizing information and tailoring safeguards to the circumstances; professional antitrust guidance limits sensitive information to people who need it to analyze the transaction. Neither source supplies a universal disclosure sequence. ^{[5] [8]}

The practical recommendation is to build restricted raw evidence and a minimized decision report as distinct outputs. Outside professionals can perform detailed work without necessarily giving the bidder's commercial personnel the underlying records. Wilson Sonsini describes an outside-adviser model in which the company receives high-level assessments rather than sensitive information directly. ^[9]

Key Changes

The changes below concern the seller's workflow. They are not a claim that a new disclosure rule took effect this year.

Change the request from a field list to a decision

Before exporting customer records, require the bidder to explain the decision the requested information will support. The Department of Justice's business-review checklist asks about an information exchange's purposes and objectives, its specificity and timeliness, and its safeguards. That checklist concerns agency business reviews; it is useful here as a question framework, not a mandatory diligence procedure. ^[10]

A proposed request should therefore state the uncertainty, the minimum evidence needed and the limitation the bidder would accept. For example, validating dependence on an economic customer group does not inherently require distributing its current price schedule. Consider whether existing nonsensitive information can answer the question before commissioning a sensitive export, consistent with Thompson Coburn's screening approach. ^[11]

Change the recipient from a title to a function

Clean teams are restricted groups operating under approved information controls. Eligibility requires examining actual responsibilities. FTC guidance excludes personnel responsible for competitive planning, pricing or strategy. A finance title alone is therefore insufficient evidence of eligibility when the person also sets prices, approves bids or directs account strategy. Counsel should evaluate those functions before access is granted. ^[5]

Identify the recipient, employer, current duties, access purpose and expected onward recipients. A transaction-specific clean-team agreement filed with the Securities and Exchange Commission (SEC) illustrates designated recipients with a justifiable need to know. It provides evidence of a contractual design choice, not a rule that other transactions can copy without analysis. ^[4]

Change the milestone from automatic access to renewed approval

A letter of intent (LOI), exclusivity or signing a purchase agreement should trigger reassessment of the request, not an assumption that sensitive information can now circulate. Mintz discusses reserving detailed information for later negotiations, while the American Bar Association emphasizes evaluating each transaction individually. The proposed process uses those principles without imposing a fixed milestone sequence. ^{[12] [8]}

Planning a future combination also needs to remain distinct from operating it. Wilson Sonsini distinguishes integration planning from implementation before closing. For this evidence pack, a diligence request should not become a channel for the bidder to direct live customer pricing, discounts or account decisions. Transaction and antitrust counsel should evaluate any request approaching that boundary. ^[9]

Implementation Considerations and Process Changes

Choose the buyer's questions before choosing fields. The seller can ask the bidder to prioritize concentration, durability, retention, margin quality and transfer conditions, rather than treating a standard customer-data request as self-justifying. SVA's customer-diligence guidance addresses one-time purchases, minimum commitments and margin variations, illustrating why a revenue ranking alone cannot answer the full commercial question. ^[13]

Table 1 is a proposed request matrix, presented as one block per buyer question. Every request remains subject to transaction-specific approval. The fields and outputs are recommendations for discussion, not permissions to disclose.

How concentrated is revenue?

• Minimum proposed fields: Stable account code, group code, period, net revenue
• Raw evidence custodian: Seller controller
• Authorized reviewer: Counsel-approved accountant
• Proposed output: Approved group shares or ranges
• Reconciliation method: Account totals to ledger and statements
• Re-identification concern: Dominant group may be recognizable
• Counsel approval: Pending, name approver
• Unresolved limitation: Ranges may limit valuation precision

Does spending recur?

• Minimum proposed fields: Consistent cohort, period sales, credit adjustments
• Raw evidence custodian: Seller finance
• Authorized reviewer: Approved financial reviewer
• Proposed output: Cohort retention and bridge
• Reconciliation method: Same cohort and accounting basis throughout
• Re-identification concern: Narrow cohort may identify an account
• Counsel approval: Pending
• Unresolved limitation: History does not prove future orders

Is margin durable?

• Minimum proposed fields: Net revenue, cost basis, rebates, freight policy
• Raw evidence custodian: Seller finance
• Authorized reviewer: Approved accountant
• Proposed output: Qualified group margin findings
• Reconciliation method: Gross-profit bridge and exception schedule
• Re-identification concern: Product mix plus margin can reveal pricing
• Counsel approval: Pending
• Unresolved limitation: Unallocated credits prevent a final conclusion

Can relationships transfer?

• Minimum proposed fields: Contract code, term, relevant consent language
• Raw evidence custodian: Seller legal
• Authorized reviewer: Transaction counsel
• Proposed output: Counsel-approved issue categories
• Reconciliation method: Contract register to customer-group map
• Re-identification concern: Unique terms may expose identity
• Counsel approval: Pending
• Unresolved limitation: Actual contracts still require legal interpretation

Is a reference necessary?

• Minimum proposed fields: Specific unresolved question and proposed contact
• Raw evidence custodian: Seller deal lead
• Authorized reviewer: Counsel-approved contact team
• Proposed output: Limited written findings if approved
• Reconciliation method: Contact scope to unresolved diligence issue
• Re-identification concern: Direct contact reveals the process and account
• Counsel approval: Separate approval required
• Unresolved limitation: Customer statements do not replace financial testing

The matrix makes trade-offs explicit. A bidder may reasonably say that ranges leave a valuation question unanswered. The response is to assess a narrower review or another evidence source, rather than silently labelling the output sufficient. FTC guidance recommends sharing the least information needed for the particular issue. ^[5]

For each request, document the following proposed decisions:

• Purpose: State what the bidder must decide and why the existing report is insufficient.
• Period: Specify the fiscal or trailing period and the treatment of subsequent adjustments.
• Population: Identify the entities, accounts and transaction types included.
• Reviewer: Record approved individuals and their actual commercial responsibilities.
• Output: Describe the permitted findings, withheld fields and qualifications.
• Expiry: Set a review point and disposition requirement under the applicable agreements.

These entries turn a broad request into a reviewable choice. Mintz recommends counsel review of a buyer's detailed information list, and Thompson Coburn recommends logging whether documents contain competitively sensitive information before ordinary data-room release. The proposed matrix operationalizes that screening without replacing counsel's judgment. ^{[12] [11]}

Reconcile the Restricted Customer File

Establish the population and financial bridge

The restricted file should preserve an intelligible path from source records to reported totals. Start with the selected period's customer subledger, reconcile it to the general ledger, then bridge the ledger to the applicable financial statements. ICAEW's financial due diligence guidance describes reconciling management information to audited financial information and addressing missing adjustments. Its analytical guidance is useful here, although it is not US law. ^[1]

Where audited statements do not exist, name the financial statements actually used and the resulting limitation. Do not label a management-account tie an audit. PCAOB AS 1105 requires accuracy and completeness work when auditors use company-produced information. That audit standard provides a useful evidence discipline by analogy; it does not turn this proposed transaction analysis into an audit engagement. ^[14]

The proposed reconciliation should include:

• Extraction controls: Save the system, query parameters, extraction time and file version.
• Population checks: Account for inactive customers, cash sales, miscellaneous accounts and excluded entities.
• Adjustment bridge: Separate posted credits, manual journals, timing items and consolidation entries.
• Cost definition: State which product, service, freight and other costs enter gross profit.
• Exception ownership: Assign unresolved differences, supporting evidence and an accountable reviewer.
• Reproduction: Retain formulas and mapping versions so another approved professional can repeat the result.

Returns and commercial adjustments deserve explicit attention. A FASB and IASB staff paper reproduces revenue guidance identifying discounts, rebates, refunds, credits and price concessions as sources of variable consideration. Deloitte's technical guidance also explains retrospectively applied volume rebates. The accountant should determine the correct treatment in the actual records, including estimates and later settlements. ^{[15] [16]}

Group economic exposure without publishing the identity key

Maintain a restricted identity key connecting legal customer names, billing accounts, ship-to records and economic-group codes. Document the evidence for each grouping, its effective period and unresolved relationships. DueDilio recommends grouping affiliates under common ownership; Deloitte reproduces the public-reporting common-control customer principle. Those sources support looking beyond billing IDs, rather than treating an accounting disclosure rule as an antitrust permission. ^{[2] [17]}

A proposed group-map record should distinguish verified ownership from management's belief that accounts share a purchasing decision maker. Central purchasing, franchise membership, a similar name or a shared address should be investigated rather than automatically treated as common control. The reviewer can preserve both legal-account and documented economic-exposure views, with exceptions visible. That is an analytical recommendation, not a new accounting rule.

Keep the key away from unauthorized recipients. The public-reporting provision reproduced by Deloitte does not itself require naming a major customer, but that narrow provision says nothing about whether a particular competitor-diligence release is lawful. The evidence pack needs its own contractual and antitrust assessment. ^{[17] [5]}

Separate Restricted Evidence from Bidder Outputs

Output A: restricted professional analysis

The restricted analysis can contain customer names, identity mappings, invoice-level prices, original contracts, cost allocations and credit support where access is approved. Its working papers should show what was inspected, the tests performed, discrepancies found and conclusions that remain unsupported. Baker Tilly recommends original-document validation of seller adjustments. A reconciled spreadsheet should therefore lead back to evidence rather than merely repeat management's assertions. ^[3]

The seller and advisers should define a permitted output before sensitive review begins. The SEC-filed clean-team agreement illustrates secure separation of restricted information and outside antitrust counsel approval before reports circulate outside the team. It is an illustrative agreement between particular parties, not a generally applicable exemption or template. ^[4]

Output B: minimized decision report

The proposed bidder report should state its period, population, grouping basis, financial tie, approved findings and limitations. It should not attach the identity key or raw account schedules by default. SourceX describes a clean-team report delivering conclusions rather than raw records, and Nixon Peabody discusses antitrust counsel review before summaries circulate. Those professional approaches support treating the report as a separate release decision. (Source: sourcex.si) ^[18]

Use a specific release checklist:

• Necessary detail: Remove fields that do not answer the approved question.
• Joinability: Compare the report with every other dataset already released.
• Cell meaning: Check whether a category or range effectively describes a single customer.
• Commercial inference: Consider what prices or account identities the recipient could infer.
• Qualification: Preserve exceptions and describe what the bidder cannot independently verify.
• Approval: Freeze the approved version and record precisely who may receive it.
• Distribution: Record delivery and the restrictions on onward sharing under the approved agreements.

Coded is not guaranteed anonymous. NIST notes that some de-identified information can be re-identified. Its subject is privacy, but the caution is useful when evaluating customer-code substitution: replacing a name does not establish that all identifying clues have disappeared. Glacier Lake similarly warns that replacing customer names with codes may be insufficient in competitor diligence. ^{[6] [19]}

The UK's Information Commissioner's Office describes linkability as a mosaic or jigsaw effect and explains that removing direct identifiers alone is insufficient. Those are technical privacy concepts, not statements of US antitrust law. In this proposed pack, a product-family report and a regional report could together expose a customer that neither names individually. Counsel should evaluate the combined release, consistent with FTC guidance on documents being pieced together. ^{[20] [5]}

No universal small-cell count or aggregation size is proposed. If the bidder already knows the market's largest account, even a broad concentration finding may reveal identity. Withholding that finding may leave genuine uncertainty. The choices include a different question, a restricted professional assessment, or a decision to leave the issue unresolved. A clean team does not make every disclosure appropriate. ^[19]

Approve Access and Handle Customer Contact Separately

Membership and technical controls

Before onboarding reviewers, counsel should assess their duties and the information involved. Wilson Sonsini identifies customer, price, cost, margin, output and future-plan information as competitive parameters. A reviewer who influences those subjects in the bidder's operating business needs careful evaluation even when assigned to a transaction workstream. ^[9]

Use access controls that implement the approved audience. Nixon Peabody discusses tightly controlled clean rooms and possible view-only settings. CISA recommends multifactor authentication (MFA) for file storage and other systems. These controls support an approved process, but they do not determine whether the underlying disclosure is appropriate. ^{[18] [21]}

Contracts, privacy and secrecy

Ask transaction counsel to review customer confidentiality provisions, permitted recipients, assignment language, consent requirements and purpose restrictions. SVA's diligence questions include contract commitments and duration; translating a contract register into a transfer conclusion requires interpretation of the actual agreements. Customer reference calls, consent requests and financial validation should consequently have separate rationales and approvals. ^[13]

Company information and personal information require distinct assessments. California's attorney general explains personal information through its connection to a person or household, and distinguishes natural-person residents from business entities. Named contacts and identifiable email addresses therefore merit a separate applicability review. This does not mean that every corporate revenue field is personal information. ^[22]

The same California guidance states that the former business-to-business personal-information exemption expired at the end of 2022. An owner should not assume B2B status resolves privacy obligations; California counsel should assess the business, information and transaction. The UK's pseudonymisation guidance is useful technical context for risk reduction, not a substitute for that US analysis. ^{[22] [23]}

The federal trade-secret definition includes conditions concerning reasonable secrecy measures and independent economic value from secrecy. These conditions support asking counsel how the proposed disclosure affects protection; they do not establish that every customer list or price is a trade secret. The applicable agreements and relevant state law still require transaction-specific review. ^[24]

Keep contact decisions distinct

Approve each of the following on its own merits:

• Financial validation: Use approved records and professional analysis to test the numbers.
• Reference calls: Specify the unresolved question, contact, script and authorized participants.
• Consent requests: Have counsel identify the actual contractual requirement and timing.
• Announcements: Decide audience, message and timing separately from diligence access.
• Operational requests: Route proposed pricing or account-direction requests to counsel before action.

This separation is a recommended control design. Permission to review a contract should not be recorded as permission to contact its customer. The underlying principle is limited use for transaction analysis, as described in the American Bar Association's guidance. ^[8]

Review the Customer-Evidence Pack (Hypothetical Example)

The records, account codes and figures in this worked distributor example are invented. Table 2 below provides the detailed arithmetic; the proposed documents described here show how the audiences differ.

Restricted version R1 would contain the identity key, ownership evidence, account-level reconciliation, credit schedule, source references and unresolved-exception record. In the invented mapping, accounts A100 and B200 both map to group G on documented parent-ownership evidence. The reviewer would reproduce each account total, confirm that ownership basis, and record the financial tie as complete while leaving credit attribution open. DueDilio's guide likewise recommends reconciling each period's customer revenue to the general ledger before analysis. ^[2]

Proposed bidder version B1 would contain only the findings counsel approves. Assume, solely for this fictional example, approval to release company totals and grouped shares to named deal-committee recipients. Its exact proposed finding is: Annual net revenue $24 million; gross profit $6 million. A verified customer group accounts for 15% of revenue and 20% of booked gross profit before allocation of $120,000 of credits already included in company totals. Group attribution is provisional; no definitive customer-margin conclusion is provided.

The report would omit the identity key, account codes, live prices and original terms. Its calculation and unresolved exception would be traceable to R1. This assumption of approval illustrates the outputs; it does not establish that such figures could be disclosed in a real transaction. Even this proposed output could be withheld or further minimized after re-identification review. Glacier Lake describes aggregated, masked or ranged counsel-approved conclusions rather than unrestricted underlying records. ^[19]

Any released numbers must use the same period, denominator and group mapping as the restricted analysis. If approved ranges replace exact shares, preserve the exact calculation in the restricted working papers and explain the reduced precision to the bidder. Do not describe B1 as an assurance report: ICAEW distinguishes financial due diligence from an audit or assurance opinion, and Baker Tilly describes quality-of-earnings work as a consulting engagement rather than an attest service. ^{[1] [3]}

Add durability tests without overstating the result

For cohort retention, retain the starting population rather than selecting only accounts that remain active. For a proposed gross revenue retention calculation, define the opening-period revenue of that cohort as the denominator, exclude new customers, and cap expansions if the stated measure excludes growth. A net retention measure should separately explain how expansions enter the numerator. These are proposed analytical definitions, not a universal B2B reporting standard.

The reviewer should also inspect purchase frequency, unusual projects, renewals, minimum commitments and subsequent credits where relevant. SVA specifically highlights significant one-time purchases and margin variations. Historical retention or an unexpired agreement should therefore be presented as evidence with limitations, rather than a promise that revenue will transfer unchanged. ^[13]

Data Analysis and Evidence

Distributor proof pack (Hypothetical Example)

All company figures, account codes and versions in this example are invented. They are original illustrative calculations, not a transaction dataset, benchmark or estimate of typical concentration. No numerical share below is a safe-disclosure threshold. The objective is a reproducible calculation, borrowing the mathematical-checking discipline described in PCAOB's definition of recalculation. ^[14]

Assume a fictional distributor's annual financial statements show $24 million of net revenue and $6 million of gross profit. The accountant has reconciled the customer extract and an explicit unallocated-credit account to those totals. Separately billed affiliates A and B belong to the same verified economic group. For this illustration only, every credit reduces revenue and gross profit equally, with no cost offset.

Table 2 shows the restricted calculation. The group subtotal is a memorandum row, not an additional amount in the total. The figures retain the same period and account mapping throughout.

The arithmetic is transparent: $3.6 million / $24 million = 15% of revenue, while $1.2 million / $6 million = 20% of gross profit. Evaluating profit concentration alongside revenue concentration follows the analytical distinction discussed by Baker Tilly. Here, however, the gross-profit share is a booked attribution with an unresolved qualification. ^[3]

The $120,000 exception is already included in company totals. Subtracting it again from total revenue would double count the adjustment. The restricted reviewer still needs evidence assigning the credits to the correct accounts and period. PCAOB's accuracy, completeness and precision principles help explain why a total that adds correctly is insufficient evidence of correct customer attribution. ^[14]

Under the invented assumption that all those credits belong to group G, group revenue would be $3.48 million and gross profit $1.08 million. Its shares would become 14.5% and 18%, respectively, with company totals unchanged. If none belongs to that group, the original shares remain. This is a sensitivity illustration under stated assumptions, not a verified allocation, probability range or economic forecast.

Implications and Future Directions

Make release approval a documented decision

A data room's existence does not describe who received a particular file. Maintain an access roster and a separate release-decision register. Mintz recommends a written log of permitted acquirer personnel, while SourceX recommends logging sample contents and recipients. The proposed register below adds version control and unresolved limitations so the approval can be reconstructed. ^[12] (Source: sourcex.si)

Table 3 is a blank release-decision register for advisers. A row should remain pending until the actual approvers decide it.

The register connects approved content to actual recipients. Glacier Lake recommends knowing which buyer individuals received each sensitive dataset. If a new recipient joins or the dataset changes, the earlier approval should be reassessed rather than presumed to cover the change. That reassessment is this report's recommended workflow. ^[19]

Pause for professional review when the evidence changes

Use explicit pause reasons rather than informal assurances:

• Role conflict: A proposed recipient sets prices or participates in account strategy.
• Recognizable cell: Combining reports exposes a customer or a live commercial term.
• Broken tie: The file no longer reconciles to the stated financial basis.
• Unresolved economics: Credits, rebates, freight or returns prevent a supported margin conclusion.
• Contract restriction: Customer terms or consent requirements remain unresolved.
• Process change: A new purpose, recipient, export or contact request exceeds the existing approval.

For a failed or withdrawn process, follow the applicable return, retention and deletion obligations, including agreed treatment of extracts, working papers, backups and legal holds. The SEC-filed agreement expressly addresses copies and extracts in its return-or-destruction provision, illustrating why deletion of the original upload alone may be insufficient under an agreement. It does not define the obligations for another deal. ^[4]

FTC personal-information guidance recommends a written retention policy and access limited to work needs. Those security principles are useful for personal data within the pack; counsel should determine the actual contractual and legal disposition requirements for the complete evidence set. Retained material should have a documented basis, custodian and access limit. ^[25]

Frequently Asked Questions (FAQs)

Can the owner disclose concentration without customer names?

Potentially, through an approved restricted review and minimized findings, but the answer depends on what the recipient can infer. NIST's re-identification caution and the professional warning about coded customer records support treating name removal as a control to evaluate, not an anonymity guarantee. Some questions may remain unanswered. ^{[6] [19]}

Does signing an LOI make customer-level pricing available?

No automatic release rule is proposed. Reassess the purpose, recipients and information at that stage with counsel. FTC guidance treats safeguards as dependent on the circumstances, and professional guidance emphasizes transaction-specific evaluation. A milestone should appear in the decision record as context, not as the approval itself. ^{[5] [8]}

Is the restricted review an independent assurance opinion?

Only the actual engagement terms can establish the professional work provided. Financial due diligence is not inherently an audit or assurance opinion. The proposed report should describe its procedures, unresolved items and limits so the bidder can decide whether further work is needed. ^[1]

Conclusion

Customer concentration due diligence for a competitor bid should begin with the buyer's questions and end with an explicit release decision. The seller's proposed pack needs a restricted source population, a financial bridge, a documented customer-group map and an exception schedule. The released report needs its own purpose, approved audience and statement of limitations.

The fictional distributor demonstrates why those components belong together. Separate billing accounts can represent shared exposure. A file can tie to the financial statements while leaving customer-level economics unresolved. A summary can preserve the same accounting logic while withholding raw evidence, but its usefulness and disclosure risk still require evaluation.

Owners should ask their accountants to establish the financial basis and their transaction and antitrust counsel to design the actual information controls. The owner then decides, with those advisers, whether the proposed evidence answers enough of the bidder's questions to justify continuing the process. Alternatives include narrowing the request, commissioning further restricted work, accepting unresolved uncertainty or declining to proceed.

The objective is credible proof with deliberate access. Coding, aggregation, a transaction milestone or a clean-team label should never substitute for reviewing what the recipient will learn. A recorded decision about a specific version, recipient list and unresolved limitation gives the seller a concrete basis for that judgment.