# Salesforce & Dynamics CRM Adoption | Adoption Layer

> Improve Salesforce and Dynamics 365 adoption and data quality: the AI reads the live screen and guides every rep field by field. No selectors, GDPR-ready.

## A CRM that stays up to date, without chasing your reps.

Opportunities with no amount, stages untouched for three months, duplicate accounts: your forecast rests on data nobody enters properly. Adoption Layer guides every rep inside Salesforce or Dynamics at the exact moment of entry, by reading the live screen.

## The CRM isn't badly designed, it's badly filled in

Your reps spend under 20% of their time in the CRM and treat it as a chore. The result: required fields bypassed with "NA", pipeline stages never updated, quotes created outside the process. Every forecast review turns into a hunt for missing data, and RevOps spends Fridays cleaning exports.

> "A rep opens a Salesforce opportunity and asks: 'I need to move this to Negotiation, what's blocking it?' Adoption Layer reads the page, spots that Weighted Amount is empty and the close date is in the past, highlights both fields, and recalls the team rule: no move to Negotiation without an attached quote."

## Guidance that adapts to your Salesforce, not the other way round

- **Live Lightning reading**: Custom objects, per-profile page layouts, conditional components: the AI interprets the DOM at the moment of the question, including after a Salesforce release or a layout change by your admin.
- **Entry rules recalled at the right moment**: Opportunity naming, required fields per stage, contact roles: your RevOps guidelines appear on the relevant field, not in a Confluence page nobody opens.
- **Pipeline friction detection**: See the stages where reps ask the most questions, the fields that get systematically bypassed, and the objects nobody understands.
- **Escalation to the Salesforce admin with context**: When the question is really a missing permission or a validation rule, the request goes to your admin or RevOps with the screen, the object involved, and the steps already tried.

## CRM workflows covered

- Creating and qualifying opportunities (amount, stage, close date, competitors)
- Converting leads and deduplicating accounts and contacts
- Generating quotes and attaching products (CPQ)
- Weekly pipeline updates ahead of the forecast review

## Why WalkMe flows break on Salesforce

WalkMe, Lemon Learning, and scripted tours rely on CSS selectors recorded at a point in time. Every Salesforce release, every page layout your admin edits, every conditional Lightning component makes them stale, and someone has to re-record them. Adoption Layer records nothing: it reads the DOM displayed at the moment of the question and maps the rep's intent to the elements actually on screen.

## Setup and privacy

Deploy the Chrome extension (Manifest V3) to the fleet via Intune, Jamf, or Google Workspace, allow-list your Salesforce or Dynamics domains, and connect your sales playbooks. No CRM changes. Customer names, amounts, and contact details are redacted in the browser before any AI inference. No cookies, no screenshots.

## Frequently asked questions

### Does it work with our custom objects and per-profile page layouts?

Yes. The AI doesn't rely on a pre-recorded map of your org: it reads the DOM as it renders for the logged-in user, with their objects, fields, and permissions.

### What happens at Salesforce releases (Spring, Summer, Winter)?

Nothing to redo. Because no selectors are stored, a renamed or moved component is simply re-read at the next question.

### Are customer data and deal amounts sent to the model?

No. Redaction happens 100% client-side: field values, amounts, names, emails, and phone numbers are replaced with tokens before inference. Only the interface structure is transmitted.

### Can we target guidelines by team or profile?

Yes. You can attach guidelines to URLs, objects, or user groups, and edit them from the dashboard without redeployment.