How to Clean a Company or Account List Before Importing Into HubSpot or Salesforce

The import is the moment you do the most damage.

A dirty account list imported into your CRM doesn't just sit there looking bad. It actively multiplies, creating duplicate company records, splitting attribution across variants of the same account, breaking territory assignments, and sending your data quality score in the wrong direction from day one.

The messy import is almost always faster to prevent than to fix. Here's how.

Where Dirty Import Lists Come From

Account lists that arrive for CRM import have a predictable set of origins, and each one brings its own flavor of mess.

Tradeshow badge scans. You badge-scanned 400 companies over three days. The export from the event app has company names in every format imaginable: "IBM Corp," "I.B.M.," "International Business Machines," "IBM," across what should be a single account in your CRM.

Partner data handoffs. A distribution partner, reseller, or agency sends you their account list in a spreadsheet. It was formatted for their internal system, not yours. Column names don't match. Account types use their taxonomy, not your CRM's.

Purchased B2B lists. A list vendor delivers a company file for ABM targeting. The data was aggregated from multiple sources with different naming conventions and different levels of completeness. Some rows have full addresses; others have just a company name and a state.

Manual outreach lists. Your sales team built a target account list from LinkedIn, company websites, and various research sources. Every rep formatted it differently. "LLC" vs. "L.L.C." vs. nothing. "New York, NY" in one cell vs. separate city and state columns.

In all four cases, the list needs work before it goes near your CRM.

The Four Problems That Cause the Most Damage Post-Import

1. Duplicate accounts under different names. "Foot Locker Inc." and "Foot Locker" and "FOOT LOCKER #5042" create three separate company records in Salesforce. Activities, deals, and contacts get split across them. Fixing this after the import is significantly more painful than preventing it before.

2. Inconsistent account types. Your CRM has a defined set of account types (Enterprise, Mid-Market, SMB, or Key Account, Regional Chain, Independent). Your import file has whatever the list vendor or tradeshow app used. "Big Box," "BB," "big-box," "Mass," and "Mass Merchant" might all be the same category in your CRM schema, or they might be completely different. Importing without normalization means running reports on garbage data.

3. Missing required fields. Your CRM has required fields for a company record: industry, billing state, phone. The import file has those fields sporadically filled. Records with missing required fields either fail the import entirely or create incomplete records that trigger data quality warnings and break automation flows.

4. Inconsistent address formatting. "123 Main St, Suite 200, Austin TX 78701" in a single cell is useless if your CRM expects street, city, state, and ZIP in separate columns. And if one row says "Texas" and another says "TX" in the state field, your state-based territory assignment rules will fire incorrectly.

The Pre-Import Cleaning Checklist

Deduplicate within the import file first. Before you even think about how the list interacts with your existing CRM data, deduplicate within the file itself. Any company that appears twice in the import, under different name formats, with different addresses, should be resolved to one row.

Standardize company naming. Pick one format per company and apply it consistently. "LLC" or not. "Inc." or not. Title case or all caps: pick one and apply it across every row in the file. This is the single highest-leverage step because CRM deduplication rules rely on name matching.

Normalize account types. Map every value in your account type column to your CRM's taxonomy before you import. If your CRM uses "Enterprise" and your file uses "Large," "Enterprise," "Corp," and "Fortune 500," all four need to map to one value before the import runs.

Parse addresses into separate columns. If addresses are jammed into a single field ("123 Main St, Austin, TX 78701"), split them into street, city, state, and ZIP before importing. If addresses are missing or incomplete, flag those rows for enrichment or manual entry rather than importing blank.

Standardize phone formatting. A consistent phone format prevents duplicate detection failures. "(512) 555-0100" and "5125550100" look like different entries to most CRMs. Canonicalize to one format before import.

Flag incomplete records separately. Don't import rows with multiple missing required fields alongside your clean records. Pull them into a separate file, work through them manually or enrich them with address/phone data, then import them as a second pass.

The CRM-Side Reality

Even a well-cleaned import file will produce some duplicates if your CRM already has records under slightly different names. Most CRMs (HubSpot, Salesforce, Pipedrive) have native duplicate detection, but their matching logic is typically exact-match or near-exact. "Foot Locker" and "Foot Locker Inc." will slip through.

This is why the pre-import file cleanup matters even if your CRM has duplicate rules. The goal is to reduce the surface area for conflicts, not eliminate every possible edge case. The more standardized your import file, the better your CRM's deduplication will perform.

When to Use a Tool vs. Clean Manually

For import files under 200 rows with mostly naming issues, manual cleaning in Excel is workable. Budget 1 to 2 hours.

For files over 200 rows with mixed account type values, address formatting issues, missing phones or ZIPs, and fuzzy naming variations across dozens of companies, manual cleaning becomes the bottleneck. It's also the part of the import workflow that gets skipped when there's deadline pressure, which is how dirty data ends up in the CRM.

ClearSheet handles account list cleanup before import. Upload the file, and the tool runs normalization across phone format, state abbreviation, account type, casing, and ZIP code; detects and flags duplicate company entries using three-pass fuzzy matching; parses addresses that are jammed into single cells; and fills missing phone or ZIP fields via Google Places (Enrich tier). Every change is shown in a line-by-line preview before you pay. First 20 fixes free.

The output is a file that's ready to import, not one that will spend the next three months accumulating merge debt inside your CRM.

FAQ

Should I clean the file before or after running my CRM's duplicate detection? Before. Your CRM's native duplicate rules work better when the input is already normalized. "Foot Locker Inc." and "Foot Locker" won't match in most CRMs even with fuzzy detection enabled, but if your file has already standardized to "Foot Locker," the CRM's rules have a better chance of catching an existing record.

My import file has people records mixed in with company data. Can I clean both at once? ClearSheet is built for company and account-level data: business names, addresses, phone numbers, account types. It's not designed for contact-level records (individual people's names and emails). For a mixed file, consider splitting it into a company sheet and a contact sheet and addressing each separately.

What format should my cleaned file be in before importing? Match your CRM's import template exactly: column names, data types, and accepted values. ClearSheet exports in the same format as your input (CSV or Excel), so if you start with the CRM's import template populated with your data, you'll get back a cleaned version of that same template.

How do I handle companies that are in my CRM under a different name than what's in the import file? After cleaning the import file for internal consistency, run a match against a CRM export before importing. Match on company name plus address or phone. Resolve any conflicts before the import runs.