No one decided that Excel would be the tool for data room reconciliation. It became the tool by default — because it's available, because analysts know it, and because for a 200-row ledger with 50 supporting documents, it works fine. The problem is that deals don't stay at 200 rows and 50 documents. And when they scale, Excel's failure modes are systematic.
What Excel-based reconciliation actually looks like
The typical Excel reconciliation setup involves three components: a copy of the seller's ledger (or a pivot of it), a manually maintained document log that tracks which files have been reviewed, and a discrepancy tab that captures mismatches and gaps. The analyst works through documents sequentially, updates the log, and records discrepancies in the tracking sheet.
This setup works at small scale. An analyst can maintain this process accurately up to roughly 500 ledger rows and 100 supporting documents before the complexity exceeds what one person can reliably track. At 5,000 rows and 400 documents, the Excel reconciliation is no longer a reliable record of what happened — it's a partially complete approximation.
The five failure modes of Excel reconciliation
1. No extraction — every match is manual
Excel doesn't extract data from PDFs. Every reference number, every amount, every date that the analyst records in the reconciliation sheet had to come from a human reading a document and typing it in. At scale, this is the bottleneck — not analysis, but data entry. Transcription errors are common and often go undetected because there's no automated verification layer.
2. No coverage guarantee
An Excel reconciliation has no mechanism for ensuring that every ledger row was actually reviewed. If an analyst skips a section, there's no flag. If a document is downloaded but never reviewed, the log shows it as complete. The reconciliation is as good as the analyst's self-discipline — which is unfair to the analyst and unreliable as a process guarantee.
3. No audit trail
Excel doesn't record when a cell was entered or changed, who changed it, or what the prior value was. In post-close disputes where the buyer's diligence process is scrutinized, an Excel file with no structured audit log is a weak evidentiary record. Reps and warranties carriers specifically ask about the diligence methodology — a spreadsheet isn't a methodology, it's a record that may or may not be complete.
In a post-close indemnification dispute, the question will be: 'What was your diligence process for verifying this category of transactions?' The answer 'our analyst reviewed them in Excel' is less defensible than 'our system produced a document-cited discrepancy report for every line item in the ledger.'
4. Merge conflicts in collaborative work
When multiple analysts work on the same reconciliation, Excel quickly becomes a merge problem. Shared drives create version conflicts. Email attachments create parallel versions. SharePoint shared editing creates race conditions where one analyst's updates overwrite another's. The result is a reconciliation file that no one is fully confident represents the current state.
5. No structured output for follow-up
Excel discrepancy logs aren't designed to feed into the deal team's Q&A workflow. Converting a discrepancy tab entry into a structured follow-up question requires manual work. Tracking which questions have been answered, which are pending, and which discrepancies have been resolved requires additional tabs that grow quickly complex. The overhead of managing the reconciliation process becomes a significant fraction of the time spent on the process itself.
What the right tool does instead
A purpose-built reconciliation tool does four things Excel can't: it extracts structured data from source documents automatically, it matches that data to the ledger deterministically with evidence citations, it produces an audit-grade record of every match and mismatch, and it surfaces discrepancies in a structured format ready for follow-up questions.
The analyst's role doesn't disappear — it shifts. Instead of spending 80% of time on data entry and matching, the analyst spends 80% of time on investigating and resolving the discrepancies that the system can't resolve automatically. That's a better use of expensive talent.
Excel was never designed for data room reconciliation. It became the default because there was no purpose-built alternative. That's no longer true.