The same read, enrich, write loop as the HubSpot version, against an org that has opinions about what it will accept.
Mechanically this is the CRM data layer again: read a filtered slice, fill the gaps, write it back, run it on a schedule. Everything true there is true here. What changes is that a mature Salesforce org has opinions encoded in it, and those opinions will reject your writes in ways that are entirely correct and completely opaque.
So the sequence is inverted. On HubSpot you build the table and then test the write. On Salesforce you test the write first, in a sandbox, with ten records, and design the table around what the org actually accepts. Teams that do it the other way round usually rebuild.
Validation rules, required fields, duplicate rules and Apex triggers all still apply to anything writing in, and a tool reporting a successful run is reporting that it sent the request. Read the error column rather than assuming a completed run also succeeded, and do it in a sandbox before anything touches production.
5 steps. The ones in blue are the data layer; everything else is plumbing you already own or can stand up in an afternoon.
Before any table exists. Which fields are required, which validation rules fire on update, which triggers run, and what the duplicate rules do. This is an hour with an admin and it determines everything after it.
Leads, contacts, accounts or opportunities missing the field you care about. Read-only, so nothing in the org changes and nothing needs approving yet.
An HTTP API column against the enrich endpoints. Identical to the HubSpot version, because this step does not care what CRM is upstream of it.
Not a hundred, not the whole slice. Ten, then read every error. This is where you find the required field nobody documented and the validation rule written in 2019 by somebody who has left.
Write to dedicated fields, promote into the trusted ones only where empty, and set the cadence from how fast the field goes stale. Then watch the API allocation for a week.
This workflow ships as a skill file: the API surface of every tool in the chain, the build order, the limits that break it, and how to check it worked. Point an agent at it and it has what it needs without reading 3 sets of vendor docs first.
--- name: salesforce-enrichment-loop description: Build a scheduled enrichment loop against Salesforce, designed around the org rules that reject writes: validation rules, required fields, duplicate r... --- # The Salesforce data layer ## What you are building ## Prerequisites ## The tools, and how to call them → Reverse Contact, Clay, Salesforce ## Build it → 6 steps ## Gotchas ## What breaks, from the vendors' own docs ## How to verify it works → 5 checks
mkdir -p .claude/skills/salesforce-enrichment-loop curl -o .claude/skills/salesforce-enrichment-loop/SKILL.md https://www.reversecontact.com/integrations-hub/skills/salesforce-data-layer.md
For Codex, Cursor, or anything else that reads a file or fetches a URL, hand it https://www.reversecontact.com/integrations-hub/skills/salesforce-data-layer.md directly. It is plain markdown with YAML frontmatter and no tool-specific syntax, so nothing in it depends on which agent is reading it.
3 of the joins in this chain have a full guide behind them, Salesforce and Clay and Salesforce and Reverse Contact among them. The chain crosses 3 of the seven layers a stack splits into, from the CRM layer outward, which is why no single vendor owns it.
Documented by one of the vendors in the chain, and easier to design around than to discover.
All four are documented by a vendor rather than observed by us, which is the useful part: a limit you can read is a limit you can design around before it costs you a week.
What each component bills on, and which line actually scales with your volume.
| Component | Billing | What drives it |
|---|---|---|
| Salesforce | Edition with API access | API access is an edition feature. Without it, none of this runs regardless of what else you hold. |
| Clay | Per action | Rows imported multiplied by enrichment columns, charged on the attempt. |
| Reverse Contact | Per credit | One call per gap row, 2 credits for a reverse lookup from an email. No dollar rate per lookup is published. |
| Admin time | Days, elapsed | Sandbox access, a change window and an hour reading validation rules. The real timeline. |
Budget the elapsed time rather than the build time. The table is an afternoon. Getting a sandbox, an admin hour and a change window is the two weeks, and no amount of tooling shortens it.
Reverse Contact resolves an email, a domain or a profile URL into a verified people and company record, with a confidence score your logic can branch on. The free trial takes no card and runs until the 500 credits a new account starts with are spent, which is enough to run this end to end before anyone signs anything.
Free trial, no card required, and 500 credits to start.
Almost always a rule somebody wrote years ago: a required field the table is not populating, a validation rule the value breaks, or a duplicate rule catching the row. Start in a sandbox, write ten records, read every error, and design around what the org accepts rather than around what you expected.
It uses part of it. Salesforce meters daily API requests by edition and licence count across the whole org, so a row-by-row write competes with every other integration you run. Batch where you can, filter the slice tightly, and watch the allocation for the first week after you schedule it.
Not mechanically. The same read, enrich, write, schedule loop. What differs is that Salesforce will reject writes that HubSpot would accept, and that getting a sandbox and an admin takes longer than building the thing. Plan for the org, not for the tooling.
Flow with an API callout can do it, and if an engineer or a capable admin owns the logic permanently that avoids a second meter. Clay earns its place when the enrichment logic is conditional and changes often, and when the person changing it does not have deploy rights in Salesforce.
Tier requirements, sync behaviour and limits read from each vendor's own documentation in September, October and November 2026. Vendors change these without notice.
Reverse Contact appears in this directory as a data provider and publishes it. Every other tool is here on merit and nobody pays for placement.