Extract contract terms to Excel from a whole folder of agreements

• Renewal planning. Finance and procurement need every auto-renewal and notice deadline in one place.
• Due diligence. A buyer asks for change of control, assignment and exclusivity terms across the customer base.
• Vendor consolidation. Operations wants to see overlapping suppliers and their termination rights.
• Regulatory or policy changes. Legal needs to know which agreements contain a data processing clause, and which do not.
• Audit preparation. Payment terms and price escalation clauses must match what finance has on record.

The quality of the spreadsheet depends on the questions you ask for each column. Write each column as a precise instruction rather than a one-word label.

• Counterparty. "Full legal name of the party other than our company."
• Effective date. "Date the agreement takes effect, as written in the document."
• Initial term. "Length of the initial term, in months or years as stated."
• Auto-renewal. "Does the agreement renew automatically? Answer yes or no, and the renewal period."
• Notice period for non-renewal. "Number of days' notice required to prevent renewal."
• Termination for convenience. "Can either party terminate without cause, and with how much notice?"
• Liability cap. "Cap on liability as stated, including any multiple of fees."
• Carve-outs from the cap. "What is excluded from the liability cap?"
• Governing law. "Which law governs the agreement."
• Payment terms. "Number of days to pay invoices."

Keep the list to what you will actually use. Ten good columns beat thirty vague ones, and each column is checked faster when its instruction is specific.

1. 1Collect the agreements in one space. Upload the files or connect the folder from Google Drive, SharePoint, OneDrive or Dropbox. Include amendments, named so they are recognisable.
2. 2Create the extraction. Choose the documents and add your columns with their instructions. ChatDocument reads each document and fills one row per agreement.
3. 3Review the table in the product. Every cell is linked to the page it came from. Click a cell and the agreement opens at the highlighted passage.
4. 4Check the cells that matter most. Start with dates, notice periods and caps on your largest agreements. These are the cells someone will act on.
5. 5Look at empty cells. An empty cell means the agreement does not contain that term, or it is expressed in a way worth reading. Open the closest passages and decide.
6. 6Export to Excel. Download XLSX or CSV. The export keeps the page reference for each cell, so the spreadsheet stays traceable after it leaves the product.

Reviewing every cell of a 200 row table is not realistic. A sampling approach works better.

• Check every cell of the five largest agreements. They carry the most risk.
• Check one column across all rows. Notice periods, for example, are quick to scan and easy to misread.
• Check every empty cell. Silence in a contract is often the finding.
• Check amendments. Make sure the extracted term reflects the amendment, not only the original agreement. If an amendment is a separate file, ask about the specific clause in both.
• Spot-check outliers. A notice period of 900 days or a cap of zero deserves a click.

Because each cell is linked to its page, each check takes seconds. That is the difference between a spreadsheet you can rely on and one you have to rebuild when someone questions it.

• Ambiguous column instructions. "Term" could mean initial term, renewal term or the whole duration. Spell it out.
• Mixed versions. Drafts and signed copies in the same folder produce two rows for one agreement. Keep the signed versions only, or name drafts clearly.
• Poor scans. Scanned agreements are read with OCR, but faint or skewed pages read less reliably. Rescan key pages if needed.
• Calculated fields. Extraction reads what the document says. A notice deadline that depends on today's date is best calculated in Excel from the extracted term and notice period.
• Forgetting the source. If you paste values into another spreadsheet, keep the page reference column.

In ChatDocument, one row of an extraction (one document) counts as one question. Each plan sets how many documents fit in one run: 10 on Starter, 50 on Team, 250 on Business and 1,000 on Enterprise. A folder of 180 vendor agreements fits in a single run on Business, or in four runs on Team. Pages count once per billing period when a document is added, and a scanned page counts as 3. See the pricing page for monthly questions and pages on each plan.

• The AI PDF to Excel feature page shows the extraction table and export.
• AI contract review covers reviewing a single agreement clause by clause.
• If a counterparty sends a new draft, compare two versions of a contract explains how to see what changed.

A spreadsheet of terms is only the start. A few ways teams use it once it exists:

• Renewal calendar. Add a column in Excel that calculates the last day to give notice from the end of the term and the notice period. Sort by that date and assign an owner to each row.
• Risk view. Filter for agreements with no liability cap, uncapped indemnities or termination for convenience by the counterparty. These are the agreements to read in full first.
• Negotiation baseline. Compare your standard positions with what was actually signed. If most agreements deviate on payment terms, the template may need updating.
• Questions for a single agreement. When one row looks unusual, open that agreement and ask follow-up questions in a thread. The answers carry their own page citations.

Keeping the spreadsheet current

Contracts change. New agreements are signed, amendments arrive, old ones expire. Instead of maintaining the spreadsheet by hand, keep the agreements in a connected folder and rerun the extraction on new documents each month. Pages count once per billing period, so re-reading documents already added in the same period does not use new pages.

Before rerunning, archive the previous export with its date, so you can see later which terms changed between two months and which agreements were added. A short note in the first row of the workbook saying which folder and which date the extraction covers saves the next reader a question.