Importing your book

For: Sellers · Cypress Managers

Prospect lists rarely arrive tidy: a dealer list from a conference, an export from someone else's CRM, a colleague's spreadsheet with whatever columns they happened to keep. CRM → Import takes that file as it is and turns it into your book — without ever creating a duplicate of someone you already know.

One template covers all three record types. Each row can produce a contact, a business, and optionally a deal, depending on which cells it carries:

• a row with an email creates or matches a contact
• a row with a business name creates or matches a business
• a row with both files the person under the company
• deal columns (title, stage, amount) open a deal attached to that row's contact or business

Rows import into your book — a seller's import never touches the platform book, and a Cypress import (or one made while acting for an org) lands where the acting-for lens points.

The four steps

1. Upload — CSV or Excel, up to 5,000 rows. Any columns, any order, headers optional; a headerless list of emails is understood as one.
2. Map columns — the mapper auto-matches your headers ("Dealership", "Main Phone", "Primary Contact Name" all land where you'd expect) and reads sample values when headers don't help — a column of email addresses maps itself. Only one thing is required: an Email or a Business Name column. Everything else is optional.
3. Review — the dry run. Nothing has been written; the screen shows exactly what committing will do, row by row and in total.
4. Import — the batch runs in the background with live progress. When it finishes, View imported contacts / businesses opens the book already filtered to what the file produced.

What the review is telling you

Every row carries up to three intent chips — one per record type:

• New — nothing in your book matches; the record will be created.
• Update — a match was found and the file fills in something the record is missing (a phone, a last name, a website). Blanks only: an import never overwrites a value someone typed.
• Match — already in your book, nothing to add. Duplicate rows inside the file collapse the same way.
• Skip — nothing to do on this side (a row without an email still imports its business; it just can't create a contact).
• Error — the row has neither an email nor a business name. Error rows are skipped at import; everything else proceeds.

Messy cells become warnings, never rejections: an unparseable phone is kept as typed, "GM: Brian" becomes Brian with a note, two people in one cell keep the first and say so. One bad cell never sinks a 2,000-row file. The Issues CSV button exports every warning and error with its row number if you'd rather fix the file and re-upload.

How matching works

• Contacts are matched by email, case-insensitively, within your book. The email is the identity — the same address can exist in another seller's book without ever colliding with yours.
• Businesses are matched by name, ignoring case, punctuation and legal suffixes — Goldie's Motors and GOLDIES MOTORS INC are one company. Rows that share a company email domain merge too: two spellings of the same dealership on babbitt-ford.com become one business with both contacts. A new business whose name merely resembles an existing one is flagged for your eyes but never merged automatically.
• State and city land on the business as its location; states are normalised ("Texas", "TX", "texas" are all TX).
• ZIP leads the address. Map a Zip column and the import resolves it to a city and state for you — the same convention every address form on the platform follows. Here the ZIP is the authority: a row whose state disagrees with its ZIP is corrected, and the correction is flagged on the row so you can see it happened. A city you supplied is kept. A ZIP that doesn't resolve is a warning, never a blocked row.
• Where the address lands depends on the row. A row with a business name describes the company's location, so ZIP, street, city and state go to the business (when it's new, or has no location on file yet). A row with no business is about the person, so the address lands on the contact — which is why contacts have one of their own.

Because matching is the same on every run, re-importing a file is safe: the second pass matches everything it created the first time and creates nothing new. An import that fails mid-way can simply be retried.

Batch options

Set on the review screen, applied to everything the import creates:

• Source label — provenance stamped on every new record (default import). This is the handle that finds them later: both the Contacts and Businesses tabs have a Source filter, and /crm?tab=contacts&source=import deep-links straight to it.
• Tags — applied to every created contact and business, useful for marking the file itself (dealer-list, west).
• Open an onboarding deal for each new business — for prospect lists: every business the import creates gets an onboarding pursuit opened at Prospecting, so the list lands as a working pipeline rather than a pile of records.

Deal columns in the file work independently of the toggle: a mapped Deal Title, Deal Stage (Prospecting, Interested or Qualified — the later stages are earned, not imported) or Deal Amount opens a deal on that row.

After the import

• Created records carry the batch's source and count as activity, so they surface at the top of the book's default sort.
• Fill-blank updates on existing records are written to the record's history, attributed to the import.
• The full batch — per-row outcome, warnings, failures — is kept and inspectable; the done screen lists any row that failed with its reason.

Vehicles are a different door: fleet spreadsheets go through Onboarding → Import, which speaks VINs, valuations and dispositions.