The most dangerous sentence in a CLM kickoff meeting is: "We'll just migrate everything."
It sounds responsible. Why leave data behind? But "migrating everything" usually means taking 20,000 unsearchable PDFs from a shared drive and dumping them into a new system. This doesn't solve your problem; it just moves your mess to a more expensive house.
The Cost of "Structure"
There is a massive difference between Storage and Intelligence.
- Storage ($0.10/doc): Uploading a PDF. You can see it, but the system doesn't know what's in it.
- OCR ($0.50/doc): The system can read the text. You can search for "Indemnification," but you get 5,000 results.
- Extraction ($5.00/doc): An AI or human identifies the "Expiration Date" and puts it in a field. Now you can run a report: "Show me all contracts expiring in Q3."
The Migration Trade-off
Why you can't (and shouldn't) extract metadata for every single legacy contract.
The "Active vs. Archive" Strategy
To avoid bankruptcy, you must triage. Do not treat all contracts equally.
1. The Active Tier (Top 20%): Contracts that are currently alive, high-value, or renewing soon.
Strategy: Full Extraction (Level 3). Spend the money here. You need alerts for these.
2. The Archive Tier (Bottom 80%): Expired contracts, NDAs from 2018, low-value POs.
Strategy: OCR Only (Level 2). Upload them so you can find them if sued, but don't waste money extracting metadata.
The "Day 2" Cleanup Myth
Many teams say, "We'll just upload them now and tag the metadata later." No, you won't.
Once the system is live, your team will be busy with new contracts. Nobody goes back to tag old NDAs. If you don't extract the data during migration, it will never happen.
The Consultant's Takeaway: Be ruthless. If a contract expired 3 years ago, do you really need to know its "Governing Law"? Probably not. Save your budget for the contracts that actually impact next quarter's revenue.