The Scenario
The Salesforce upload window is at 2 PM. You found out at 9 AM that the 2,200-row account list you're importing has never been through any kind of address validation — it was compiled from three different sources over eight months, each with its own formatting conventions, and whoever merged them together didn't bother standardizing anything. Street addresses that say "Ave." in one block and "Avenue" in the next. Zip codes with four digits instead of five. A "State" column where someone typed "California" and someone else typed "CA" and a third entry just says "Cal."
Your CRM admin warned you last import that Salesforce's duplicate detection runs on standardized address fields. If the addresses don't match their canonical format, records that should merge will create duplicates instead.
The bad version:
- Open the workbook, filter for obvious formatting issues, and fix them row by row until your eyes blur
- Export the addresses to an online validation tool with a 500-row limit per run, handle the four separate exports and four sets of results, and manually reconcile each batch back into the master workbook
- Try to invent a HIGH/MEDIUM/LOW scoring scheme on the fly using nested IF formulas across five columns, realize the logic doesn't cover UNCONFIRMED addresses that have a partially valid component, and spend 45 minutes patching edge cases
Nobody hired you to spend the morning on address hygiene. The Salesforce admin is asking for an ETA. The account manager whose accounts these are has a pitch at 3 PM and needs the CRM up to date before it.
The Easy Way: One Prompt in SheetXAI
SheetXAI is an AI agent that lives inside your Excel workbook. It reads the structure, calls Google Address Validation for each row, and writes the standardized addresses and quality scores directly into your columns — no batch exporting, no formula engineering, no manual reconciliation.
Go through each row in my Excel sheet, validate the address assembled from columns B through E using Google Address Validation, write the corrected address to column F, and add a score of HIGH, MEDIUM, or LOW to column G based on how complete and deliverable the address is
What You Get
- Column F filled with standardized addresses in Google's canonical format — abbreviations consistent, capitalization normalized, zip codes corrected where the API can confirm the right value
- Column G showing HIGH for DELIVERABLE addresses with complete components, MEDIUM for UNCONFIRMED addresses where the API returned partial data, and LOW for UNDELIVERABLE or incomplete addresses
- The original columns B through E left untouched so you can diff the before and after
- Any row where the API returned an error flagged explicitly in column G rather than silently given a LOW score
What If the Data Is Not Quite Ready
The address comes from a single free-text column instead of separate fields
One of the three source files mashed everything into one column — "123 Main St, Chicago, IL 60601" — without splitting it into components.
For each row in my Excel sheet where column B has a single free-text address, parse and validate it using Google Address Validation, write the standardized full address to column F, and score it HIGH, MEDIUM, or LOW in column G based on the validation verdict
International addresses are mixed in with US addresses
About 300 rows are Canadian accounts that someone added to the same workbook, and they need to be validated against Canadian postal conventions instead of US ones.
Validate the address assembled from columns B through E for each row in my Excel sheet — for rows where column H says "Canada," pass the address to Google Address Validation with country set to CA instead of US, write the standardized address to column F, and score it HIGH, MEDIUM, or LOW in column G
Some rows have a zip code but no city or state
A subset of accounts only recorded the zip code, and the city and state columns are blank.
For each row in my Excel sheet where the City and State columns are blank but the Zip column is populated, attempt to infer the city and state from the zip code before validating with Google Address Validation, then write the inferred city and state into those columns, the full standardized address to column F, and a note in column G indicating the city and state were inferred
Standardize, score, remove LOW-quality rows, and write a final import-ready output in one pass
Before handing off to the CRM admin, you want to produce a clean version of the workbook with all LOW-quality addresses removed and a summary count written to a header row.
In my Excel sheet, validate every address assembled from columns B through E using Google Address Validation, write the standardized address to column F and the quality score (HIGH, MEDIUM, or LOW) to column G, then delete all rows scored LOW, and write the total count of HIGH, MEDIUM, and removed rows into cells A1, B1, and C1
The pattern is to ask for the validation, the scoring, and the cleanup in a single prompt — you get the final import-ready file without an intermediate step.
Try It
Get the 7-day free trial of SheetXAI and open your account workbook — the one going into Salesforce — then ask it to validate, standardize, and quality-score every address before the upload window. Hub: How to Connect Google Address Validation to Excel. Related: Bulk Validate Shipping Addresses in an Excel workbook.