Inside a Business Lead CSV: Sample File and Field Guide
Read a business lead CSV as a set of business records with contact details attached. Before importing it, check what each row represents, which columns you selected and how multiple emails or people are associated. Those checks prevent a correctly formatted file from creating the wrong CRM records.
This guide reflects Leadz's main export options reviewed on September 23, 2026. Download the illustrative business-only sample CSV to practice: it contains 16 business columns with fictional records, not all 20 currently available export columns.
Understand the row and choose the columns
One exported row represents a business record, which may describe a location. It is not automatically a unique corporate account or a named person. Two locations can share a website or central inbox, while one business row can contain several email addresses and associated people.
The main Leadz export dialog offers 20 columns, with 14 selected by default. Business Name is required. The six initially unselected columns are Additional Categories, Phone Count, Email Count, Zip Code, Rating and Reviews. Review the selection each time; defaults are a starting point for the task, not a requirement to keep every column.
The teaching download contains the 16 business-level columns and omits the four person-profile columns. Its fictional records demonstrate shared inboxes and location differences for import practice. Use a product export to inspect the full selection of columns and your actual contact data.
For account research, business name, category, website and geography may be sufficient. For a contact handoff, include the relevant email and associated-person fields. Avoid exporting a large file solely because more columns are available; every field should have a destination or a review purpose.
CSV export requires an eligible paid plan. Plan access and available credits can affect newly unlocked exports, so check the product's confirmation and current pricing details. Keep a copy of the selected columns and source list with your working file.
Use the field dictionary to plan the mapping
These are the CSV headers, which can differ slightly from labels in the export dialog. In particular, the option labeled Complete Email Profiles produces the header Email Profiles JSON.
CSV header | What the value means |
|---|---|
Business Name | Recorded business title; not necessarily its legal entity name |
Category | Primary category describing the business record |
Additional Categories | Other recorded categories, combined in one cell |
Primary Phone | Main recorded phone value, potentially a branch or shared line |
All Phones | Recorded phone values combined in a semicolon-separated cell |
Phone Count | Count of stored phone entries, with a primary-phone fallback |
Primary Email | Main recorded email value |
All Emails | Recorded email values combined in a semicolon-separated cell |
Email Count | Count of stored email entries, with a primary-email fallback |
Contact Names by Email | Available person names prefixed by their associated email |
Contact Roles by Email | Available reported roles and companies, grouped by email |
Contact LinkedIn by Email | Available LinkedIn profile URLs grouped by email |
Email Profiles JSON | Structured email-to-profile information within one CSV cell |
Website | Recorded website value for further business research |
Address | Recorded business address |
City | Recorded city value |
State | Recorded state value, which may need normalization |
Zip Code | Postal code; preserve it as text |
Rating | Recorded rating where a value is emitted |
Reviews | Recorded review count where a value is emitted |
Use Phone Count and Email Count to inspect the contact entries available for review. Research staff size or purchasing responsibility separately. Rating and Reviews describe the recorded customer-review information; keep those fields separate from your own qualification score.
Inspect Primary Email alongside All Emails. Their formatting and available values can differ; do not assume one is a perfect copy of the other. Keep the raw columns before choosing a preferred address in your working sheet.
For definitions beyond file structure, the availability-versus-verification guide explains what an email on file establishes and which checks belong before outreach.
Preserve the email-to-person relationship
Person columns retain the address associated with each name, reported role or profile. Read them one email at a time: several profiles can exist in a business record, while other addresses in the same row have no person information.
Consider an invented business with two addresses: office@cedar.example and maya@cedar.example. Only the second has a saved person profile. The human-readable cells might look like this:
Column | Illustrative value |
|---|---|
All Emails | office@cedar.example; maya@cedar.example |
Contact Names by Email | maya@cedar.example: Maya Chen |
Contact Roles by Email | maya@cedar.example: Operations Manager at Cedar Studio |
Contact LinkedIn by Email | Blank in this example |
Maya's name belongs to maya@cedar.example. It should not be copied onto the office inbox. A blank LinkedIn cell does not weaken the email association; it simply leaves that particular field unknown.
Human-readable values for multiple people are separated with semicolons. Where one profile reports several current roles, those role descriptions can be separated with a slash. Keep the email prefix when transferring values into a research sheet so the relationship remains visible.
Email Profiles JSON retains structured information for workflows that can parse it. Treat it as a JSON value inside a CSV cell, not as a ready-made CRM job-title column. Parse the CSV first and then parse that cell's JSON. When there are no saved profile rows, the structured value can be an empty array while the human-readable person fields are blank.
A reported role is evidence for further review. Confirm whether the person is relevant to your offer and the account you are approaching, especially when branches share contact details. The business database guide explains how to keep companies, locations and people connected without collapsing them into one object.
Keep blanks, zeros and text values distinct
A blank cell means no value was emitted for that field. It does not prove the business has no website, phone number or contact person elsewhere. Record new research in separate working fields rather than silently filling the source columns.
Rating and Reviews deserve particular care: the current formatter can leave zero-valued ratings or review counts blank. A blank review cell therefore should not be interpreted as a confirmed zero. Contact-count fields have their own fallback rules and should not be treated as a substitute for inspecting the addresses themselves.
Microsoft's guidance on preserving leading zeros explains how Excel's numeric conversion can change imported values and how to choose text types during import. Changing the display format later cannot recover digits already lost.
The illustrative download prefixes phone values with an apostrophe to keep them as text in spreadsheets. Preserve the raw file and inspect the import preview rather than stripping punctuation from every column. Any cleanup rule should have a defined purpose and a way to compare the result with the original.
Let the CSV parser handle commas and quotation marks
CSV is a file format, not simply a line of text that can always be split at every comma. A company name, address or structured profile can itself contain a comma. Quotation marks tell the parser when such content belongs inside one field.
RFC 4180, published in October 2005, documents common CSV formatting: fields containing commas, line breaks or quotation marks are enclosed in double quotes, and a quotation mark inside such a field is doubled. It is an informational format reference, not evidence about contact quality.
For example, the fictional name Cedar, Plumbing should remain one cell when represented as "Cedar, Plumbing". A quotation mark inside a quoted company name will appear doubled in the raw file. A spreadsheet importer normally handles those characters without manual replacement.
Use your spreadsheet's CSV import command, check the delimiter and preview several rows. Do not switch the file delimiter to semicolon just because All Emails contains semicolons: those separate values inside a cell, while commas separate CSV columns. For automated work, use a CSV parser rather than string splitting.
If the preview shows shifted addresses, extra columns or profile text scattered across cells, stop before importing. Reopen the untouched export and correct the import settings; manually moving cells can hide a parsing error that affects many other records.
Map accounts and contacts before importing a batch
Create a mapping that states which destination object receives each column. Business Name belongs to a company or account field. A person name belongs to a contact only after its email association is understood. A location address should remain connected to the correct branch.
Read your CRM's current import rules before choosing identifiers. For example, HubSpot's import documentation, checked on September 23, 2026, describes unique identifiers for updating records and making associations, including Record ID and supported email or company-domain identifiers. Choose the appropriate identifier for the object rather than assuming a business name will prevent duplicates.
Be careful with a shared company domain. Two branches may belong to one corporate account while still needing separate location records. If that distinction matters, establish it in your destination model before loading the file. The deduplication workflow covers the review decisions.
Test a small tagged batch containing a multi-location business, an empty contact field and a row with multiple email values. Confirm that existing records update as intended, blank imports do not erase researched details, and contact associations remain correct.
Reconcile imported, rejected and intentionally excluded records with the source batch. Save the mapping and any rejection reasons. A second upload should follow an understood correction, not serve as an experiment that creates another set of possible duplicates.
Keep the export separate from your working database
An export is a dated input to the workflow. Add account owner, source date, research notes, suppression status and next action in your working system; those fields are not part of the 20-column export described here.
Keep the original file unchanged, name the working version clearly and document any normalization or merges. Before a campaign handoff, check the selected accounts against existing customers, open opportunities and exclusions. A freshly downloaded record should not undo a previous opt-out or duplicate decision.
Start by importing the business-only teaching sample, then inspect a small real export from the category that fits your audience. Confirm the column choices and contact relationships before increasing volume. The goal is a file your next teammate can understand and use without reconstructing the research.