How to Audit Your HubSpot Imports Before They Wreck Your Database

Sixty-seven percent of data quality issues in HubSpot databases originate from CSV imports. Not from form submissions, not from API integrations, not from manual data entry. From imports — the blunt instrument that every RevOps team reaches for when they need to get data into HubSpot quickly, whether that data is ready or not.

The problem is not that imports are inherently dangerous. The problem is that they bypass every quality gate your CRM has. A form submission runs through field validation, deduplication rules, and workflow triggers. An API integration can enforce schema constraints and reject malformed records. A CSV import does almost none of that. It takes whatever you give it, maps it to whatever columns you tell it to, and writes it directly into your contact database — misspelled company names, invalid emails, duplicate records, and all.

This post is a pre-import audit checklist. Follow it before every import and you will prevent the damage instead of remediating it after the fact.

Why HubSpot's Built-In Import Validation Is Not Enough

HubSpot's import tool does perform some validation. It checks that your CSV columns map to existing HubSpot properties. It validates that date fields contain date-formatted values. It flags rows where a required property is missing. It warns you about potential duplicates based on email address matching.

This is format validation, not accuracy validation. And the distinction matters enormously.

Format validation asks: "Is this value shaped like an email address?" It checks for an @ symbol, a domain, a TLD. The address jsmith@compnay.com passes format validation perfectly. It is also invalid — a typo that will hard-bounce on first send and chip away at your sender reputation.

Accuracy validation asks: "Does this email address actually exist, accept mail, and belong to the person in this record?" HubSpot's import tool does not do this. It cannot, because accuracy validation requires real-time queries against mail servers, DNS records, and verification APIs that operate outside HubSpot's import pipeline.

The same gap exists across every field type:

• Job titles: HubSpot accepts "VP of Stuff" and "CEO/Founder/Janitor" without comment. No standardization, no normalization against your existing title taxonomy.
• Phone numbers: Any string of digits passes. No country code validation, no format normalization, no verification that the number is dialable.
• Company names: "Salesforce", "salesforce.com", "Salesforce, Inc.", and "SFDC" all import as four distinct companies. HubSpot does not reconcile them.
• Country fields: "US", "USA", "United States", "United States of America", and "America" are five different values in a free-text field.

The Pre-Import Audit Checklist: Five Steps

Every CSV import should pass through these five checks before it touches your HubSpot database. The total time investment is 30 minutes to 2 hours depending on file size. The remediation cost of skipping them is measured in days.

Step 1: Email Verification

Run every email address in the import file through a verification service before import. This is non-negotiable for any import containing more than 100 records.

Email verification services (ZeroBounce, NeverBounce, BriteVerify, Hunter.io) check each address against three layers:

1. Syntax validation — is the address properly formatted?
2. Domain validation — does the domain exist and have active MX records?
3. Mailbox validation — does the specific mailbox exist and accept mail?

The output categorizes each address as valid, invalid, risky (catch-all domains, role-based addresses like info@ or sales@), or unknown. Your import rule should be simple: import valid addresses only. Flag risky addresses for manual review. Exclude invalid and unknown addresses entirely.

The cost of verification is typically $0.003-0.008 per address. The cost of importing 500 invalid addresses and sending to them — degraded sender reputation, reduced inbox placement across your entire database, manual cleanup — is orders of magnitude higher.

Step 2: Field Standardization

Before import, normalize every field value against your existing HubSpot property standards. This means:

Job titles: Map incoming titles to your standardized seniority and function taxonomy. If your HubSpot instance uses "VP of Marketing" as the canonical form, then "Vice President, Marketing", "VP Marketing", and "Vice President of Marketing & Communications" should all be normalized before import. Build a mapping table once and reuse it.

Company names: Reconcile against your existing company records. If "Salesforce" already exists in your database, "salesforce.com" and "Salesforce, Inc." in your import file should map to the existing record, not create new ones.

Geographic fields: Standardize country, state, and city values to ISO formats or your chosen canonical forms. "US" and "United States" must resolve to the same value.

Phone numbers: Normalize to E.164 format (+1XXXXXXXXXX for US numbers). Strip extensions, parentheses, dashes, and spaces. Validate country codes.

Industry fields: Map to your existing industry taxonomy. If your HubSpot uses "Software" as the industry value, "SaaS", "Technology", "Software/Tech", and "Computer Software" need to resolve to "Software" before import.

Step 3: Duplicate Pre-Check

Before importing, identify which records in your file already exist in HubSpot. HubSpot's import deduplication uses email address as the primary match key — which means it misses every duplicate that has a different email address but is clearly the same person.

Run a pre-check that matches on multiple fields:

• Email match: exact email addresses that already exist in HubSpot
• Name + company match: same first name, last name, and company name with a different email — likely a job change or secondary address
• Domain + name match: same email domain and similar name — possible duplicate with email variation

For each match, decide before import: update the existing record, skip the import record, or merge fields from both. Making this decision during import prep is straightforward. Making it after 5,000 duplicate pairs have been created is a multi-day cleanup project.

Step 4: Required Field Validation

Define your minimum viable record — the set of fields that must be populated for a contact record to be useful in your workflows. For most B2B operations, this includes:

• Email address (verified)
• First name
• Last name
• Company name
• Job title or seniority level

Any record in your import file that does not meet this minimum should be flagged. You have three options for incomplete records:

1. Enrich before import: Use enrichment tools to fill missing fields before the data enters HubSpot. This is the highest-quality option.
2. Import to a staging list: Import incomplete records to a dedicated list that triggers an enrichment workflow in HubSpot. This keeps them out of your active segments until they are complete.
3. Exclude from import: If the record lacks an email address or company name, it may not be worth importing at all.

Step 5: Test Import on a Subset

Never import the full file on the first pass. Take a random 5% sample (or 100 records, whichever is larger) and import it first. Then check:

• Did field mapping work correctly? Are values landing in the right properties?
• Did deduplication rules fire as expected?
• Did any workflows trigger that should not have (welcome emails, lead assignment, lifecycle stage changes)?
• Are formatted values (dates, phone numbers, currencies) displaying correctly?
• Did any records create new companies when they should have associated with existing ones?

A test import catches mapping errors, workflow conflicts, and format issues on 100 records instead of 10,000. The five minutes it takes is the cheapest quality assurance investment in your entire import process.

The Three Import Disasters Every RevOps Team Has Seen

Disaster 1: The Tradeshow Lead Dump

Your team attends a conference and returns with 3,000 badge scans exported as a CSV. The sales team wants them in HubSpot by Monday. The file contains: first name, last name, email, company, and nothing else. Half the email addresses are personal accounts (gmail.com, yahoo.com) because attendees used personal badges. Fifteen percent of the company names are abbreviated or misspelled because they were manually keyed by booth staff.

What goes wrong: 3,000 records import with no job title, no seniority, no industry, and no phone number. Duplicates with existing contacts are created because the badge-scan email differs from the existing HubSpot record email. Workflows trigger welcome emails to personal addresses. Lead scoring assigns default values because every enrichment field is empty. The sales team gets 3,000 "new leads" that are operationally useless.

What should happen: Verify emails (expect 15-20% invalid from tradeshow sources). Enrich before import to fill job title, seniority, company size, and industry. Run duplicate pre-check against name + company. Import to a dedicated tradeshow list with a separate nurture workflow, not your main marketing pipeline.

Disaster 2: The Purchased List

Someone on the team buys a list of 10,000 "targeted contacts" from a list broker. The list arrives as a CSV with 15 columns that do not match your HubSpot property names. Half the records are three to five years old. The list contains contacts from industries and company sizes that are outside your ICP.

What goes wrong: 10,000 records import with incorrect field mapping because column names did not match. Invalid emails hard-bounce on first send, pushing bounce rate above 2%. Spam trap addresses in the list trigger blacklisting. Contacts outside your ICP pollute your segments and inflate your contact tier billing. Cleanup takes weeks.

What should happen: Verify every email address (reject the 25-40% that are invalid). Filter by ICP criteria before import — remove contacts outside your target company size, industry, and geography. Standardize field names to match your HubSpot properties. Import to a quarantine list with a 30-day engagement window before promoting to active segments.

Disaster 3: The CRM Migration Import

You are migrating from Salesforce, Pipedrive, or another CRM to HubSpot. The export file contains 50,000+ records spanning five years of accumulation, including leads that never converted, contacts that bounced years ago, companies that no longer exist, and custom field values that have no equivalent in HubSpot.

What goes wrong: Every record from the old CRM imports into HubSpot, including the garbage. Custom field values land in free-text fields without validation. Historical engagement data is lost or improperly mapped. Duplicates multiply because the old CRM's deduplication rules were different. Your "new start" in HubSpot begins with the same data quality problems that plagued your old CRM.

What should happen: Audit the source database before export. Suppress contacts that have not engaged in 18+ months. Verify all email addresses. Map custom fields to HubSpot equivalents with explicit transformation rules. Import in batches by segment (active customers first, then prospects, then historical) so you can validate each batch before proceeding.

The Post-Import Audit Workflow

Even with a thorough pre-import audit, you need a post-import verification step. Run these checks within 24 hours of every import:

Bounce Rate Monitoring (First 24 Hours)

If you send any email campaign to imported contacts within the first week, monitor bounce rates in real time. A hard bounce rate above 1% on an import segment indicates that your pre-import email verification either was not thorough enough or was not performed at all. Stop sending to the import segment immediately if bounces exceed 2%.

Deduplication Sweep (Within 48 Hours)

Run a full duplicate analysis within 48 hours of import. Even with a pre-import duplicate check, edge cases slip through — contacts whose email changed between your pre-check and the import, records that matched on email but had field conflicts that created partial duplicates, and company records that were auto-created by the import rather than associating with existing companies.

HubSpot's native deduplication catches exact-match email duplicates. For fuzzy matches (same person, different email), you need either a manual review or a tool that does probabilistic matching on name + company + domain.

Enrichment Timing (Within 72 Hours)

If you imported records with missing fields (which you should have identified in Step 4), trigger enrichment within 72 hours. The longer incomplete records sit in your database, the more likely they are to be included in campaigns and sequences where missing data will degrade personalization and targeting.

Priority order for post-import enrichment:

1. Records missing job title or seniority (blocks lead scoring and routing)
2. Records missing company size or industry (blocks segmentation and ICP filtering)
3. Records missing phone number (blocks outbound calling sequences)

Workflow Audit (Within 1 Week)

Check every active workflow that uses contact properties populated by the import. Look for:

• Workflows that triggered on import records when they should not have (e.g., welcome emails sent to contacts who are existing customers in the import)
• Workflows that did not trigger because imported values do not match enrollment criteria (e.g., lifecycle stage values formatted differently than what the workflow expects)
• Lead scoring anomalies — contacts scored higher or lower than expected because imported fields have different values than your scoring rules anticipate

Import Quality Metrics to Track

Build a simple dashboard that tracks import quality over time. Four metrics matter:

Track these per import source. You will quickly identify which sources consistently produce clean data and which consistently require heavy remediation. A tradeshow badge-scan export and a product-qualified lead export from your own application should not receive the same level of pre-import scrutiny — the latter is likely clean, the former almost certainly is not.

Building an Import Quality Culture

The import checklist above is a process. Processes work when they are enforced, and they are abandoned when they are not. Three organizational changes make pre-import auditing stick:

Centralize import permissions. Not everyone on the team should be able to import contacts into HubSpot. Limit import access to RevOps or a designated data steward. This is not about control — it is about ensuring that every import passes through the audit checklist before it touches the database.

Create an import request form. When the sales team returns from a conference with leads, or marketing purchases a list, the request should go through a standard intake process: source of data, number of records, intended use, and timeline. The data steward then performs the pre-import audit and either approves the import or sends back a remediation list.

Track import quality by source. When you know that Conference X's badge scans consistently have a 30% invalid email rate, you can build that into your timeline and budget accordingly. When you know that your product's signup form consistently produces 98% valid records, you can fast-track those imports through a lighter audit process.

The goal is not to slow down your team's ability to act on new data. The goal is to ensure that every record entering your database meets a minimum quality standard that protects your deliverability, your segmentation accuracy, and your reporting integrity.

How MarketingSoda Refine Handles Import Quality

MarketingSoda Refine includes an import quality scoring engine that automates the five-step checklist described above. Before any import touches your HubSpot database, Refine scores the file across email validity, field completeness, standardization compliance, duplicate risk, and freshness indicators.

Records that pass all quality gates import directly. Records that fail specific checks are routed to remediation workflows — enrichment for missing fields, verification for suspect emails, deduplication for matched records. Nothing enters your production database without meeting your quality threshold.

The result is that every import maintains your database quality instead of degrading it. Your sender reputation stays protected, your duplicate rate stays controlled, and your database health metrics trend in the right direction with every new data source you add.

Join the MarketingSoda Refine waitlist to get import quality scoring built into your HubSpot data operations from day one.