Import CSV. Leads & Deals

When to Use CSV Import

CSV import is your best ally in three situations:

• Migrating from another tool: You're leaving a CRM or competing tracking tool and want to recover your history
• Importing historical data: You have months (or even years) of data in Excel or Google Sheets files
• Unsupported CRM: Metrikia natively integrates 6 CRMs. If yours isn't one of them, CSV import is the solution

Pro tip: Even if your CRM is natively supported, CSV import is still useful for loading historical data from before the integration connection date.

Lead Import

Download the Template

Go to Settings > Import > Leads and click Download CSV Template. The file contains the expected columns with examples.

Required Fields

A lead must have at minimum one of these identifiers:

• Email (recommended, best deduplication accuracy)
• Phone (international or French national format accepted)
• First name + Last name (if neither email nor phone is available)

Optional Fields

Upload Process

1. Drag and drop your CSV file (or click to browse)
2. Metrikia displays a preview of the first 5 rows
3. The column mapping interface opens (see dedicated section)
4. Validate the mapping and launch the import
5. A progress bar shows real-time advancement
6. At the end, an import report summarizes: rows imported, updated, skipped, errors

Deal Import

Required Fields

Optional Fields

Pro tip: Always import your leads BEFORE your deals. A deal must be linked to an existing lead. If the lead referenced by email doesn't exist, the row will be flagged as an error.

Deduplication Logic

This is the most powerful feature of CSV import. Metrikia doesn't create duplicates, it recognizes existing leads through PII hashing (Personally Identifiable Information).

How It Works

Metrikia uses HashedPii to compare imported data with existing leads:

1. Email hash (score: 1.0) : The email is normalized (lowercase, Gmail dot removal, +alias handling) then hashed. This is the most reliable identifier.
2. Phone hash (score: 0.8) : The number is normalized to E.164 format for France: +33 prefix, leading 0 stripped. So 0612345678, +33612345678, and 06 12 34 56 78 are all recognized as the same number.
3. Name hash (score: 0.6) : The name is normalized via ICU NFD transliterator: accents are removed, case is harmonized. Jean-François and jean francois match.

Match Scoring

Metrikia calculates a match score (LeadMatchScore) based on available hashes:

• Email only: score 1.0 → Match
• Phone only: score 0.8 → Match
• Name only: score 0.6 → No match (below threshold)
• Phone + Name: score 1.4 → Match

The match threshold is 0.8. Above it, the existing lead is updated. Below it, a new lead is created.

Column Mapping

Automatic Detection

Metrikia automatically recognizes common column names:

Manual Mapping

For unrecognized columns, the mapping interface allows you to:

• Associate each column with a Metrikia field
• Create a custom field on the fly
• Ignore a column (don't import it)

Best Practices

1. Test with 10 rows first. Import a small sample to verify that mapping and deduplication work correctly before launching the full import.
2. Include email for best accuracy. Email is the most reliable identifier for deduplication (score 1.0). Without email, Metrikia must rely on phone or name, which are less precise.
3. Use ISO 8601 date format (YYYY-MM-DD). Ambiguous formats like 01/02/2025 (January or February?) cause errors.
4. Monetary values in euros with decimals (1500.00) or in cents (150000). Metrikia detects and converts automatically.
5. UTF-8 encoding required. If your file comes from Excel, export as "CSV UTF-8" to preserve accents and special characters.

Troubleshooting

Common Errors

Pro tip: After a large import, check the import report in Settings > Import > History. You'll find the row-by-row detail of creations, updates, and errors. It's the best way to verify your data is clean.