Basics

How does a CRM work?

A plain walkthrough of how a CRM works from records to views to automation, using SchemaDesk as a concrete schema-first example.

Updated September 22, 2026

We built SchemaDesk. These guides are our honest take on the CRM category, including where we are not the right fit.

The basic flow

You define what you track, enter or import records, update them as work happens, and use views and reports to steer the team. Automation can react to changes later.

In SchemaDesk, that flow starts with objects and fields in a workspace, then records, relations, and saved views on top.

Data layer

Objects define record types. Fields define typed attributes. Relations connect records. Permissions decide who can read or change them.

Work layer

People live in list views and record pages. They filter, sort, search, and update stages. Meetings should run from those views, not from private copies.

Create or import a record
Assign an owner
Move a stage or status
Log the next step

Improvement layer

Once the data is trusted, automation and reporting amplify it. Without trust, those layers amplify noise.

How SchemaDesk works in practice

SchemaDesk resolves workspace membership from the session, stores business data as workspace-scoped records, and renders UI from field types. That is how a CRM works when customization is a first-class requirement.

FAQ

Does a CRM work offline?

Some offer limited offline modes. Most teams treat the cloud workspace as the live system and use mobile or desktop clients against it.

How does import fit the daily flow?

Import is how you bootstrap or refresh bulk data. Daily work afterward is usually manual updates, integrations, or forms writing into records.

How do views work in SchemaDesk?

SchemaDesk views save filters, sorts, and columns so each team can look at the same records through the lens they need.