Skip to main content

How do I get data from a Google Sheet my team edits daily into HubSpot or Salesforce reliably?

Don’t push every edit. Give each row a permanent ID, run a sync every 5–15 minutes that sends only changed rows, and upsert them into the CRM against that ID. At Smplfy we build this for $1,500–$6,000 in two to four weeks. Price depends on how many columns need mapping rules and which direction the data flows.

Why sheet-triggered syncs break

Most teams start with a Zapier or Make scenario that fires on a new or updated row. That works for a week and then starts to drift. The tool isn’t the problem. The problem is that a shared sheet changes in ways a row trigger can’t tell apart.

  • Edited rows. Change one cell in an old row and some setups ignore it, while others re-send the whole row as a new record. Updated-row triggers usually watch a single column you choose, so edits anywhere else never sync.
  • Renamed or moved columns. Mappings are tied to header names or positions. When someone renames Phone to Mobile or inserts a column, values land in the wrong CRM field or go blank. Nothing errors, so nobody notices.
  • Partial saves. Google Sheets saves each cell as it’s typed. A trigger can fire once the first cell is filled, sending a contact with a name and no email, and then the rest of the row arrives as a second event.
  • Duplicates. Sorting, filtering or inserting rows changes row numbers. A sync that tracks rows by position, or matches on an email with a typo, creates a second contact or deal.
  • Deletes. Deleting a row sends no event. The CRM record stays, and the sheet and CRM stop agreeing.
  • Bulk changes. Pasting 400 rows or running a find-and-replace can trigger hundreds of calls at once. That hits HubSpot or Salesforce API limits and some rows fail without anyone seeing it.

Google Apps Script’s on-edit triggers have a related gap. They fire for edits a person makes by hand, but not for changes made through the API or by other scripts, so data imported into the sheet some other way never syncs.

How a reliable sheet-to-CRM sync works

The fix is to stop reacting to edits and instead compare what the sheet holds now with what you last sent. Five parts do the work.

  1. A permanent row ID. One column holds an ID that is assigned once and never changes, no matter how the sheet is sorted. The CRM stores the same value: a unique property in HubSpot or an External ID field in Salesforce.
  2. A header check before every run. The job confirms every expected column is present by name. If one was renamed or deleted, the run stops and alerts a named owner instead of writing bad data.
  3. A ready signal. A Status column set to Ready, or a rule that a row must be unchanged for 10 minutes, means half-typed rows never sync.
  4. Change detection. The job saves a fingerprint (hash) of each row it has synced. On each run it sends only rows whose fingerprint changed, in batches that stay inside the CRM’s rate limits.
  5. Upsert, not create. Each record is written against its row ID. A matching record is updated and a missing one is created, so running the sync twice does no harm. Failed rows are logged with the CRM’s error message and retried, and their status shows up in a column the team can see.

Worked example

Six people on a sales-ops team work a 3,000-row lead sheet, and about 150 rows change each day. The goal is Google Sheet → HubSpot contact and company → deal created when Stage reaches Qualified. A scheduled job runs every 10 minutes and sends about 15 changed rows per run in one batch call. It writes a Synced at timestamp and any error text back into two columns on the sheet.

At our $85 an hour, this takes about 35–50 hours: roughly $3,000–$4,250. Most of that time goes on mapping rules (picklist values, owner names to HubSpot user IDs, date formats) and on error handling, not on the API calls. A one-way sync with ten plain text fields is near $1,500. A two-way sync, where CRM changes flow back to the sheet, moves toward $6,000.

Your options compared

Option Upfront cost Monthly cost Who maintains it What breaks first
Manual CSV import None Staff time Whoever imports Duplicates when the matching column is wrong
Zapier or Make row trigger A few hours to set up Typically $20–$100 (market range) Whoever built it Edited rows, renamed columns, partial saves
Off-the-shelf sync connector Setup only Subscription, varies by vendor Vendor plus your admin Custom fields and conditional logic it can’t express
Custom scheduled sync (Smplfy) $1,500–$6,000 Hosting, usually a few dollars Your engineer or ours, using a runbook Upstream CRM field changes, caught by alerts
Move off the sheet to a form or internal tool From $4,000 (Smplfy) Hosting Your engineer or ours Little at sheet level; the data is validated when entered

We built a Stripe-to-Salesforce and Google Sheets finance sync on this pattern, and it has run for two years with minimal maintenance.

When to move off the sheet

A reliable sync makes the sheet safe to keep for a while. It doesn’t make a spreadsheet a system of record. Move the data into a form, a CRM view or a small internal tool when two or more of these are true:

  • More than five people edit the sheet every day.
  • Rows trigger money or commitments: invoices, contracts, payouts.
  • You’ve added protected ranges, dropdowns and warning colours just to keep the sync working.
  • You need to know who changed a value and when, and version history isn’t enough.
  • The sheet has passed about 10,000 rows, or several tabs are copies of each other.
  • The team mainly edits the sheet to fix what the CRM got wrong.

For Hoopes Vision we replaced a spreadsheet workflow with a database pipeline and reporting dashboard, which ended the manual monthly reports. That’s the usual endpoint once a sheet is doing a database’s job.

When not to hire anyone

If the sheet changes weekly, fewer than about 50 rows move, and a duplicate is a minor annoyance, use the CRM’s own import with a unique matching column, or a Zapier zap with a Ready column. Pay for a custom sync when bad data would cost you a deal, an invoice or a week of cleanup.

Frequently asked questions

Why does Zapier create duplicate contacts from my Google Sheet?
Usually because the zap creates a record on every trigger instead of finding and updating an existing one, or because it matches on a value that changes, such as an email with a typo. Edits to older rows can also be re-sent as new rows. Add a permanent ID column, store it on the CRM record, and update against that ID.

How often should a Google Sheet sync to HubSpot or Salesforce?
Every 5–15 minutes suits most daily-edited sheets. It’s fast enough that the CRM feels current and slow enough that half-typed rows are finished before they sync. Real-time per-edit syncing adds failure points and rarely helps.

What happens to the CRM when someone deletes a row in the sheet?
Nothing, unless the sync is built to notice. Deletes send no event, so a reliable sync compares the current list of row IDs with the last run and flags the missing ones. Most teams archive or tag those CRM records rather than deleting them automatically.

How much does a custom Google Sheets to Salesforce sync cost?
Smplfy builds them for $1,500–$6,000 at $85 an hour, in two to four weeks. A one-way sync with simple fields sits at the low end. Two-way syncs, picklist and owner mapping, and creating related records such as deals or opportunities push it higher.

Can I keep using the Google Sheet after the sync is built?
Yes. The sync is designed so the team keeps working in the sheet, with a status column showing what synced and what failed. Once the sheet starts triggering invoices or contracts, or needs a real audit trail, plan to move that data into a form or internal tool.

Related answers

Want a scoped estimate for your sheet and CRM?

We’ll send a free written scope or have a 20-minute call. Have ready: which CRM you use, roughly how many rows and editors the sheet has, which columns matter, and what someone fixes by hand today. Start a conversation.