CSV Merge Field Mapping for Cold Email, Done Right

If your CSV exports clean, stable headers, importing it into any sender is just field mapping. The convention is one column per step (like step_1_subject and step_1_body through step_N_subject and step_N_body), plus the standard contact columns: first_name, last_name, company, email and title. This guide shows the exact mapping for Smartlead, Instantly, Apollo, and HubSpot, one section each. Bookmark it for your next campaign.

First, understand the column layout

A personalized sequence CSV looks like this. Each row is one contact, and each step's subject and body live in their own columns, already written for that person.

Because the writing is already done per row, your sending tool does not generate anything. It just inserts the right column into the right step. That is what the merge field does. A SimpleSend export can also carry an optional step_N_delay_days column with each step's wait, any extra columns you mapped on the way in, and a talk_track column if you generated call cards. Map what you need and ignore the rest.

Smartlead

In Smartlead, create a new campaign and upload the CSV as your lead list. Map email, first_name, last_name and company to Smartlead's standard fields, and import the step columns as custom fields. When you build your sequence, add one email step per step in your CSV, and reference the matching columns with double curly braces.

• Email 1 subject: {{step_1_subject}}
• Email 1 body: {{step_1_body}}
• Email 2 subject: {{step_2_subject}}, and so on.

Send a test to yourself first to confirm the variables resolve. If you see the raw {{step_1_body}} text in the test, the column name in your CSV does not match the variable exactly, so check casing and spelling.

Instantly

The flow in Instantly is nearly identical. Upload the CSV when creating a campaign, and Instantly maps known columns and keeps the rest as custom variables. During import you can confirm or adjust how each column maps. Build your sequence steps and drop the same variables in, using Instantly's variable syntax, again one step per pair of columns.

One tip: Instantly lets you preview a few contacts before launching. Use it. Scan three or four rows to make sure each contact's body matches their company, not someone else's, which catches any column misalignment instantly.

Apollo

In Apollo, import the CSV into a list, and the per-step columns come in as custom fields on each contact. When you build a sequence, reference those custom fields as variables in the subject and body of each step. Apollo's variable picker will show your imported fields, so you select step_1_subject for the first email's subject line and so on.

Since Apollo also bundles data and sending, this is a common path for teams already living in Apollo: enrich and source there, draft the emails elsewhere, then bring the finished copy back in as fields.

HubSpot

HubSpot treats the columns as contact properties rather than campaign variables, so there is one extra setup step. Before importing, create custom contact properties that match your column names, such as a single-line text property called step_1_subject. Then import the CSV and map each column to its property. (SimpleSend's own HubSpot connection is read-only: it imports lists but never writes back, so this import is a manual step.)

In a HubSpot sequence, you reference these as personalization tokens. Insert the token for step_1_body into the first email, and HubSpot fills in each contact's stored value. Use sequences, which send one to one from your connected inbox, rather than marketing email: HubSpot's acceptable use policy doesn't allow marketing email to purchased lists or contacts who haven't opted in. The mapping is a little more manual than the dedicated senders, but it works the same way underneath.

A quick troubleshooting checklist

Sending tools referenced: Smartlead, Instantly, Apollo, and HubSpot. Exact menu labels change as these tools update, so use this as a map rather than a literal click path.