A practical, vendor-neutral checklist to use when comparing ERP/accounting platforms. Use these questions in demos, RFPs, and reference calls to surface risk early.

Questions that surface hidden spreadsheet dependency

Use these questions early in demos to separate real process change from spreadsheet work that simply moves out of sight.

• Which spreadsheets disappear on day 1, by name and by owner?

• Which spreadsheets still survive after go-live, and why do they remain?

• If a spreadsheet remains, is it temporary, required, or just not yet handled in the product?

• What changes after approval force re-review, updated sign-off, or a repost?

• Can one board or investor metric be traced from the final report back to the ledger and source records without Excel?

• How are exceptions resolved without off-system workarounds such as side calculations, email approvals, or manual trackers?

• What evidence of that process can a reviewer inspect in the system versus outside it?

Implementation & time to value

Core questions

• What does a typical implementation plan look like (phases, owners, dependencies)?

• What is the expected time to first meaningful value (e.g., first close, first consolidated reporting) vs. full rollout?

• What internal resources are required (finance, IT, operations), and how many hours per week should each expect to spend?

High-signal follow-up questions (timeline, support, migration, pricing)

• What is included in the quoted implementation fee vs. billed separately (project management, training, integrations, data migration, custom reports)?

• What are the most common reasons implementations slip, and what controls do you use to mitigate them (scope control, change orders, decision logs)?

• How do you handle data migration (what years/periods, what objects), and who is responsible for validation and reconciliation before go-live?

• What does user acceptance testing (UAT) look like, and what are the explicit go/no-go criteria for launch?

• What post–go-live “hypercare” support is provided (duration, hours, response SLAs), and what does ongoing support cost after that period?

• If we need to roll back or pause, what is the contingency plan and how is business continuity maintained?

Integrations & data quality

Core questions

• Which systems do you integrate with out of the box (banks, AP/expense, CRM, payroll, billing, data warehouse/BI), and which require custom work?

• How does data sync work (real-time vs. batch), and how are failures detected, retried, and surfaced to users?

• How do you validate and monitor data quality end-to-end (completeness checks, duplicate detection, reconciliation reports)?

High-signal follow-up questions (security, governance, migration)

• Do you provide a documented API (and/or webhooks)? What are the key limitations (rate limits, pagination, long-running jobs), and what is your deprecation policy?

• How is access secured for integrations (OAuth/SAML/SCIM, service accounts, IP allowlists), and what compliance attestations do you provide (e.g., SOC 2)?

• How do you handle master data management (chart of accounts, customers/vendors, products, entities) and prevent key mismatches across systems?

• What built-in reconciliations are available (subledger-to-GL, bank-to-cash, billing-to-revenue), and can we schedule them with alerts?

• Can we export all raw and transformed data at any time (including mappings), and in what formats?

• How do you support ongoing schema/mapping changes without breaking downstream reporting (versioning, change logs, approvals)?

Accuracy & explainability

Core questions

• How are key numbers calculated (e.g., allocations, accruals, consolidations, revenue recognition), and where are those rules configured?

• Can a reviewer trace any reported number back to source transactions and transformations (full audit trail)?

• How are exceptions identified, reviewed, and corrected—and how do those corrections flow through to downstream reports?

High-signal follow-up questions (controls, auditability, pricing)

• What accuracy or reconciliation metrics do you track, and how are they measured during implementation and ongoing operations?

• When upstream data changes (late invoices, corrections, reclasses), how are recalculations handled (versioning, effective dates), and can we reproduce prior-period results?

• Are calculations deterministic and explainable to auditors (rule logic, inputs, intermediate steps), and can we export that evidence?

• What controls exist around configuration changes (role-based access, separation of duties, approvals, change history)?

• What is the process for resolving discrepancies with the vendor (support workflow, escalation path, SLAs), and what is included vs. paid add-on support?

• How is pricing affected by complexity drivers that impact accuracy (entities, transaction volume, modules, custom rules), and what usage limits should we plan for?