Free · No account · No upload
The Data Readiness Scorecard
Before an AI agent reads your client data, find out what it would find. Export a CSV from your CRM, drop it here, and get a 0–100 readiness score across the five things that decide whether agents help or embarrass you: completeness, duplicates, formatting, household linkage, and notes quality.
Drop your CRM export here
A CSV with one row per client or per account. We recognize common column names from Wealthbox, Redtail, Salesforce, and plain spreadsheets — headers are matched case-insensitively and common aliases are accepted.
No file loaded yet. The sample is a fictional 28-client firm with realistic defects planted in it.
Overall score, 0–100
–
–
Five dimensions
Weighted: completeness 30% · duplicates 20% · format consistency 20% · household linkage 20% · notes quality 10%. Scores are computed per unique client where the dimension is about clients, and per account row where it is about accounts.
What we found
Next steps
This is exactly the cleanup agentdatasync performs before connecting agents — deduplication, household linking, and field standardization, with your CCO’s sign-off on the result. How it works · The Audit tier · Docs
Expected columns
One row per client, or one row per account (client fields repeat across their accounts — the Scorecard deduplicates clients by client_id, or by name + date of birth when no ID column exists). Missing columns don’t stop the run: we tell you which are missing and score what’s possible, marking the affected dimensions.
| Column | Used for |
|---|---|
client_id | Identifying unique clients |
first_name, last_name | Duplicate detection |
email | Completeness, duplicate detection |
phone | Completeness, format consistency |
dob | Completeness, duplicate detection, format checks |
state | Completeness, valid 2-letter code check |
household_id, household_name | Household linkage |
account_number, account_type | Account linkage (rows with an account but no household) |
notes | Notes quality (substantive vs. stub) |
Aliases accepted include, among others: client id, e-mail, mobile, date of birth / birthdate, household, account #, and note / comments.
How to read the grade
- 85–100 — Agent-ready. Minor cleanup only. Agents can read this data without routinely tripping over gaps.
- 60–84 — Close. Usable bones, specific defects. Targeted cleanup — usually duplicates and household links — gets you ready.
- 0–59 — Not yet. Agents on this data would answer from duplicates, blanks, and unlinked accounts. Clean first; it is cheaper than apologizing to clients later.
A note on honesty: this Scorecard checks the shape of your data — presence, format, linkage — not its truth. A present-but-wrong phone number scores as present. It is a readiness screen, not an audit, and it makes no compliance determination of any kind.