Reconcile is a file-first reconciliation tool. You upload two files — CSV, Excel, or JSON — and Reconcile matches them, flags every discrepancy, and produces clean, audit-ready reports — a PDF summary and an Excel workbook of every row. No ERP connection required. No IT involvement. No onboarding call. Just upload your files and get your reports.

Reconcile is built for small finance operators — freelance bookkeepers, small accounting firms, startup finance teams, e-commerce operators reconciling payout reports, and agencies reconciling ad spend. What they share is not their industry. It is their situation: two files, a recurring problem, and no tool that meets them where they are.

Any situation where two files should agree but don't. Bank statement vs. ledger. Stripe payout export vs. accounting records. Invoice list vs. payment records. ERP export vs. bank export. Warehouse count vs. ERP stock. If you have two files and need to know what matches, what doesn't, and why — that is Reconcile's job.

Three things Excel cannot do. First, it does not break when your column names change — Reconcile recognizes what a transaction reference looks like regardless of what it is called in a given export. Second, it produces a structured report you can hand to a client or auditor as is, not a highlighted spreadsheet you still have to clean up and explain. Third, it tells you exactly what matched, what did not match, and why — with no formula to doubt and no manual review to second-guess.

Generic CSV comparison tools are format-blind. They require your files to be identically structured and produce a raw diff — a list of differences with no financial context. Reconcile understands financial file patterns. It handles messy, inconsistent exports automatically, surfaces likely column mappings and a likely match key from your data, and produces a structured reconciliation report — not a diff. The output is something you can actually use, not something you still have to interpret.

No. Reconcile works alongside whatever system you already use. It does not connect to your accounting system or replace any part of it. It works with the files those systems produce. You export, you upload, you get your reports.

Yes. The reconciliation engine works on any structured data with a shared identifier. Inventory counts, customer records, order lists — if you have two files and a column to match on, Reconcile can compare them. That said, the product is purpose-built for financial reconciliation workflows and that is where it performs best.

No. You upload your files, choose and map the relevant columns, confirm the match key, set the tolerances that fit the compared fields, and run. The setup stays short. No formulas, no scripts, no configuration project.

Yes, on the Teams plan. Teams includes a shared workspace where multiple users can access, run, and review reconciliations together. The pay-as-you-go plan is for individual use.

Yes. Reconciliation logic can be saved and reused. Once you have configured a run — column mapping, match key, compared fields, and tolerances — you can reapply the same setup the next month without starting from scratch. Upload new files, review the saved setup, and run.

Yes. The PDF report is structured for submission, and the Excel workbook provides the traceability behind it — matched rows, mismatches, missing rows, and duplicates clearly separated and labeled. Together they are designed to hold up in an audit or client review without additional formatting or explanation.

No. That is the point. Reconcile is built for files as they actually come out of your systems — inconsistent column names, mixed date formats, amounts formatted differently across sources, missing rows, extra headers. You upload your files as they are. Reconcile handles the rest before the comparison even starts.

CSV, Excel (XLSX), and JSON. Files up to 10 million rows per file are supported across all plans.

You need at least one field that both files share — a transaction reference, an invoice number, an order ID, anything that can identify the same row in both files. That becomes the match key. After upload, you map the equivalent fields, confirm that key, and set tolerances where they are needed on the compared columns.

Reconcile handles this as a starting point. The normalization engine analyzes your files before the reconciliation runs and detects which columns likely represent the same thing, even when they are named differently. It then surfaces a suggestion like "We think `reference` in file A matches `memo_ref` in file B." You confirm or adjust that mapping, then move on to the match key and tolerances.

Yes. Sample files are available directly in the upload section so you can run a reconciliation and see the output before uploading your own data.

Yes. Use the sample files provided, or create a small anonymized version of your own data. Your first two runs are free, so you can test the full workflow — including the PDF report and the Excel results workbook — before committing to anything.

A good match key uniquely identifies each row in both files. Transaction references, invoice numbers, and order IDs are ideal. Amount-only matching is weak — the same amount can appear multiple times and produce false matches. Reconcile will warn you if the suggested match key has low uniqueness, meaning it is likely to produce unreliable results.

You can combine two columns to create a composite key — for example, matching on date plus amount together. Reconcile supports multi-column match keys. If no reliable key exists, the results will reflect that uncertainty and you will need to review matches manually.

Upload your files. Reconcile profiles them silently. Choose and map the relevant columns. Confirm the match key and set the tolerances that fit the compared fields. Run. Review the results in the UI. Download your PDF report and Excel results workbook. The whole process from upload to report typically takes under two minutes for well-structured files.

Go to the upload section on the home page. Upload your two files. No account is needed to start — you will be prompted to create one before your run completes so Reconcile knows where to send your reports and how to count your free runs.

Reconcile matches rows based on the match key you confirm — a column or combination of columns that both files share. If the value in the match key column is identical across both files after normalization, the rows are considered a match. Reconcile then compares the additional columns you selected and flags any differences.

Yes. Composite keys — matching on two or more columns together — are supported. For example, you can match on item code plus branch, or date plus reference number. This is useful when no single column is unique enough on its own.

A matched record is a row that exists in both files with the same match key value. Once matched, Reconcile compares the columns you selected and reports whether they agree or differ.

An unmatched record is a row that exists in one file but has no corresponding row in the other file based on the match key. Unmatched records are reported separately so you can investigate them directly.

The extra rows in the larger file are reported as unmatched. Reconcile does not assume missing rows are errors — it reports them clearly and lets you decide what they mean in your context.

If the same match key value appears more than once in either file, Reconcile flags it. Duplicates are reported separately because they represent an ambiguity — Reconcile cannot reliably determine which row in one file corresponds to which row in the other when the key is not unique.

Rows with null or blank values in the match key column are reported as unmatched. Reconcile does not attempt to match on empty values because there is no reliable basis for comparison.

No. Reconcile matches by key value, not by row position. Your files do not need to be sorted or ordered in any particular way.

Yes. Once you have configured a run, the column mapping, match key, compared fields, and tolerances can be saved and applied to future runs with new files. This is particularly useful for recurring monthly workflows where the file structure stays consistent.

Yes. You can modify the column mapping, match key, compared fields, or tolerances and rerun without uploading your files again, as long as the files are still within their retention window.

A short set of decisions is yours: confirm or adjust the column mapping, confirm the match key, and set per-column tolerances for the compared fields. Reconcile handles normalization, matching, discrepancy detection, and report generation after that.

Yes. Results are displayed in full in the UI before you export. You can review matched rows, unmatched rows, and discrepancies, and decide whether the results are ready to export or whether you want to adjust the configuration and rerun.

Yes. The current limit is 10 million rows per file across all plans. This covers the vast majority of real-world reconciliation files including large ERP exports and high-volume payout reports.

Yes. The current file size limit is 50MB per file. If your file exceeds this, exporting it as CSV rather than Excel typically reduces the size significantly.

Pay as you go supports 2 files per run. The Teams plan supports up to 5 files per run. The Business plan supports up to 50 files per run, enabling multi-entity reconciliation across large file sets in a single operation.

Yes. There is no restriction on running reconciliations in parallel. Each run is processed independently.

Yes. Storage limits vary by plan. Older reconciliations beyond the retention window are automatically removed. You can export your results at any time to keep a permanent copy outside of Reconcile.

The free runs have the same limits as paid runs — up to 10 million rows per file, the PDF report, the Excel results workbook, and the audit dashboard included. The free runs are not a reduced version of the product.

Yes. Moving from pay as you go to Teams increases the files per run from 2 to 5. Moving to Business increases files per run to 50 and removes the row limit per file. Contact us for Business plan details.

The upload is discarded and no credit is consumed. You can retry immediately. If upload failures persist, check your file format and size, and contact support if the issue continues.

No tool that is honest will make that guarantee. What Reconcile does guarantee is that the output is transparent — every matched row, every unmatched row, and every discrepancy is clearly labeled and traceable. You can see exactly what the engine did and why. If something looks wrong, you will see it. That is a stronger assurance than a guarantee that hides its work.

By requiring a unique match key rather than matching on amounts or dates alone. Amount-only matching produces false matches when the same value appears multiple times. Reconcile warns you when the suggested match key has low uniqueness and recommends adding a second column to strengthen it.

Through normalization. Many apparent mismatches in manual reconciliation are not real differences — they are formatting inconsistencies. A trailing space, a different date format, an amount written as "(4,200.00)" in one file and "-4200" in another. Reconcile normalizes these before comparison so formatting differences do not produce false discrepancies.

Reconcile applies configurable tolerance to numeric comparisons, and separate window-based tolerances for date, datetime, and time fields where relevant. Small rounding differences — for example, $0.01 mismatches from fee calculations — can be flagged separately from material discrepancies so you are not chasing noise.

The normalization engine standardizes date formats before comparison. "26-Apr", "2026-04-26", and "Apr 26 2026" are all recognized as the same date and compared correctly regardless of which format each file uses.

The match key was not unique enough, producing false matches. One file contained duplicates that were not expected. A column had mixed data types that normalization could not resolve. The wrong fields were mapped or selected for comparison. Or a tolerance was too wide or too narrow for the data. In most cases, reviewing the mapping, match key, comparison columns, and tolerances resolves the issue.

Start with the column mapping and the match key. Check whether the fields were aligned correctly and whether the match key is truly unique in both files — if the same value appears more than once, results will be unreliable. Then check whether the comparison columns and tolerances match the data you expected. Most unexpected results trace back to one of those areas.

Yes. Given the same files and the same configuration, Reconcile will always produce the same output. There is no randomness in the matching or comparison logic.

Reconcile reports discrepancies as differences — it does not classify them as errors or timing differences automatically, because that classification requires business context that only you have. The report gives you the information to make that judgment: what the value was in each file, and whether the row was matched or unmatched.

First check the column mapping, match key, comparison columns, and tolerances as described above. If the configuration looks correct and the results still seem wrong, contact support with your run ID. We will investigate without requesting your actual files.

Every run produces two files. First, a PDF report — the general report of the run, with the summary, the discrepancy breakdown, and the audit trail, ready to send to a client or auditor. Second, an Excel workbook with every row from both files sorted into five sheets: matched, mismatches, missing in source, missing in comparison, and duplicates.

Matched contains the rows that agree across both files, with their comparison results. Mismatches contains the rows that matched on the key but differ on a compared value, with both values shown. Missing in source and missing in comparison contain the rows that exist in only one file. Duplicates contains the rows that appear more than once. Every row is labeled, so the recipient can understand the results without needing access to Reconcile.

The PDF report gives you the high-level picture — how many rows matched, how many did not, and a breakdown of discrepancies by type — in a format suitable for client delivery and audit submission. The Excel workbook gives you the row-level detail — every row with its status and comparison result — suitable for further analysis, follow-up, and audit review.

Yes. You can download the PDF report and the Excel workbook as many times as you need within the retention window of the run. Each download reflects the same fixed snapshot of the results.

Yes. Both the PDF report and the Excel workbook capture the results at the point of the run. If you rerun with different settings, the new run produces a new snapshot. Past exports are not affected.

Yes. The PDF report is structured for submission — the summary, the discrepancy breakdown, and the audit trail in one document. The Excel workbook provides the traceability behind it — every row has a clear status and every discrepancy is documented. Together they are designed to hold up in an audit without additional formatting or explanation.

Yes. You can select which columns from your original files to include, in addition to the reconciliation status columns that Reconcile adds automatically. The five-sheet structure itself is fixed so that every export is consistent and traceable.

The PDF report and the Excel workbook are available for download within the retention window of your plan. After that window, the run and its exports are removed. Download your files promptly if you need to keep them permanently.

You have two options. Pay as you go — buy credits and use them when you need them. One run costs $12, with discounts for larger packs down to $6 per run for 40 runs. Credits never expire. Or choose the Teams plan at $149 per month, which includes 60 runs per month, up to 5 users, a shared workspace, and priority support. Your first two runs are always free, no credit card required.

A run is one complete reconciliation — from uploading your files to receiving your reports and audit dashboard. A run covers up to 10 million rows per file and includes everything: normalization, matching, discrepancy detection, the on-screen results, the PDF report, and the Excel results workbook.

Your first two runs are completely free. Both runs include the full product — up to 10 million rows per file, the PDF report, the Excel results workbook, and the audit dashboard. No credit card is required to use your free runs.

After your two free runs are used, you need credits or a Teams plan to run additional reconciliations. Viewing past results and managing your workspace does not consume credits.

A single run costs $12. Credit packs offer volume discounts: 5 runs for $49 (roughly $10 per run), 15 runs for $119 (roughly $8 per run), and 40 runs for $249 (roughly $6 per run).

Yes. The larger the pack, the lower the cost per run. The 40-run pack at $249 brings the cost down to $6.25 per run — less than the cost of one hour of bookkeeping time.

No. Credits never expire. Buy a pack and use them at your own pace.

Unused credits can be refunded within 30 days of purchase. Credits that have been consumed cannot be refunded.

On the Teams plan, additional runs beyond the 60 included are charged at $2.49 per run. You will not be cut off — you will simply see the additional charge on your next invoice.

Yes. The cost of a run is shown before you confirm and start. You will never be charged without seeing the cost first.

Yes. On the Teams plan, credits and the monthly run allowance are shared across all users in the workspace.

The Teams plan at $149 per month includes 60 runs per month, up to 5 users, up to 5 files per run, a shared workspace, full reports (PDF and Excel workbook), the audit dashboard, and priority support.

Pay as you go is for individual users who reconcile occasionally and prefer not to commit to a monthly fee. Teams is for bookkeepers and small finance teams who reconcile regularly across multiple clients or entities and need shared access and a predictable monthly cost. If you are running more than 30 reconciliations per month, Teams is better value.

Yes. You can upgrade to Teams at any time. Unused credits from a pay-as-you-go pack remain in your account and can be used after you switch.

Yes. Reconcile is designed for individual operators and freelancers. You can pay as an individual with a personal card. No company account or VAT number is required.

Prices shown are exclusive of applicable taxes. Tax will be calculated and shown at checkout based on your location.

Yes. Reconcile reserves the right to adjust pricing. Existing credit packs are honored at the price paid. Teams subscribers will be notified in advance of any pricing changes.

Reconcile processes the structure and values of your files — column names, row values, and data types — for the purpose of normalization and matching. We process only what is necessary to run your reconciliation. We do not read, analyze, or store the content of your files for any other purpose.

Your files are processed for comparison only and deleted immediately after your run completes. We never store them, read their content for any purpose other than matching rows, or share them with anyone. What is inside your files is your business, not ours.

No. Uploaded files are deleted immediately after processing. Reconciliation results — the matched rows, unmatched rows, and discrepancies — are stored temporarily within your retention window so you can review and export them. The original files are never retained.

For comparison only. Reconcile's engine reads column names and row values to perform normalization and matching. It does not interpret the meaning of your data, extract information from it, or use it for any purpose beyond running your reconciliation.

No. Your uploaded data is never used to train models, improve algorithms, or for any purpose beyond running your specific reconciliation. It is deleted immediately after your run.

No. Your data is never shared with third parties. Reconcile does not serve advertising and does not sell or share customer data with any external party.

Access to customer data is strictly limited. No Reconcile staff member can access the content of your uploaded files — they are deleted before any human review could occur. Reconciliation results are accessible only to support staff when investigating a reported issue, and only with your explicit request.

Yes. Each workspace is fully isolated. There is no shared data infrastructure between customers at the application level.

Yes. You can delete individual reconciliation results at any time. You can also request full account deletion, which removes all associated data permanently.

No. Payment is handled by a third-party payment processor. Reconcile never stores or has access to your card details.

All data transmitted between your browser and Reconcile is encrypted using TLS. This applies to file uploads, API calls, and all other communication with the platform.

Reconciliation results are encrypted at rest using AES-256 encryption. Uploaded files are deleted immediately after processing and are never stored in an unencrypted state.

Yes. Files remain encrypted throughout the processing pipeline. Decryption occurs only within the processing environment and only for the duration of the normalization and matching operation.

Workspace isolation is enforced at the application and infrastructure level. Every request is authenticated and scoped to a specific workspace. It is not possible for one customer's session to access another customer's data.

Accounts are protected by password authentication with bcrypt hashing. Two-factor authentication is available and recommended. Login attempts are rate-limited to prevent brute force attacks.

Yes. Two-factor authentication is available for all accounts and can be enabled in your account settings.

Yes. Every reconciliation run is logged with a timestamp, the configuration used, and the identity of the user who ran it. This audit trail is available to workspace administrators and can be included in compliance documentation.

Reconcile is working toward SOC 2 Type II certification. In the meantime, our security practices align with SOC 2 principles across availability, confidentiality, and security. Contact us if you have specific compliance requirements.

In the event of a security incident affecting customer data, affected customers will be notified promptly with details of what occurred, what data was involved, and what steps are being taken. We follow responsible disclosure practices.

Reconcile processes data for comparison only and deletes files immediately after your run. If your files contain personal data covered by GDPR or similar regulations, you remain the data controller and Reconcile acts as a data processor. Review our Data Processing Agreement for details.

You can reach support by email at support@reconcile.app or through the in-app help widget. Include your run ID if your question relates to a specific reconciliation.

Pay-as-you-go users can expect a response within 48 hours on business days. Teams plan users receive priority support with a target response time of 24 hours. Business plan customers receive dedicated support with an SLA.

No. Your uploaded files are deleted immediately after your run and cannot be retrieved by support staff. If you need to share data to troubleshoot an issue, share a small anonymized sample directly. We will never ask for your original files.

Service status is published at status.reconcile.app. Planned maintenance and unplanned outages are communicated via email to registered users and posted to the status page in real time.

Yes. Contact us through the website and we will schedule a walkthrough. That said, the product is designed to be self-explanatory — your first two runs are free and require no setup, so the fastest way to evaluate Reconcile is to run your own files through it.

Yes, integrations are on the roadmap for the Business tier. The self-serve product is deliberately file-first — no integrations required. For Business customers with specific integration needs, contact us to discuss your requirements.

Yes. Scheduled reconciliation — automatically running a saved configuration when new files arrive — is planned. This will be particularly useful for bookkeepers with recurring monthly workflows.

Yes. API access is planned for the Business tier, allowing teams to trigger reconciliation runs programmatically and retrieve results without using the web interface.

Yes. Fuzzy matching — matching on similar but not identical values — and ML-based match suggestions are on the roadmap. The current engine handles normalization and exact matching after standardization. Fuzzy matching will extend coverage to cases where data quality issues cannot be fully resolved by normalization alone.

Contact us directly at feedback@reconcile.app or use the feedback link in the app. We read every message and the roadmap is shaped primarily by what paying users ask for most consistently.

Your use of Reconcile is governed by the Terms of Service and Privacy Policy, both available on the Reconcile website. By using the product, you agree to these terms.

You do. Reconcile does not claim any ownership over your uploaded files or reconciliation results. You retain full ownership of your data at all times.

For the data you upload — your reconciliation files — Reconcile acts as a data processor. You are the data controller. You determine the purpose and means of processing. Reconcile processes the data only according to your instructions, which are expressed through your reconciliation configuration.

Reconcile is designed to support compliance with GDPR for users in the European Union and EEA. We are working toward additional certifications. If you have specific regulatory requirements, contact us before using the product with sensitive data.

No. Reconcile produces a reconciliation report that supports your internal controls and audit procedures — it does not replace them. The output is a tool for your review, not a substitute for professional judgment or established compliance processes.

In the event of a shutdown or acquisition, registered users will be notified in advance with sufficient time to export their reconciliation results. All customer data will be handled in accordance with the Privacy Policy and applicable law.

Reconcile's Terms of Service are governed by the laws of the jurisdiction specified in the Terms of Service document available on the website. Please review the current Terms of Service for the applicable jurisdiction.