Pre-flight checks, anti-pattern audits, deployment runbooks, and post-incident reviews. Each checklist is short enough to run in a single sitting and specific enough that finding even one match pays back the read.
Get new checklists in your inbox
Free. No spam. Unsubscribe anytime.
Browse by type
The Collection
Every checklist is a 15-minute walk-through. Read on page or download the PDF. Free, no gate.
Salesforce will charge Flex Credits for successful agent calls over MCP and API. The rate is not published. Three questions to ask before metering starts.
Salesforce ended support for Workflow Rules and Process Builder in 2025, but they still run. Rank by risk, drop some, rebuild the rest as flows you approve.
The Salesforce sharing and security model on one page: OWD, the 6-rung record-access ladder, restriction vs scoping rules, FLS and the Apex/Flow bug. Free PDF.
Vendor-neutral 50-point checklist for evaluating any AI agent, copilot, or Agentforce vendor before it touches production Salesforce. Free PDF.
The 47-point checklist to run before you launch Agentforce: data quality, process clarity, permission hygiene and automation overlap. Skim in 15 minutes.
The 38-point checklist for auditing Salesforce permissions yourself: profile sprawl, integration users, permission set hygiene and roles. Skim in 20 minutes.
A 15-minute audit checklist of the 15 most common Flow anti-patterns in production Salesforce orgs. Signal, why it hurts, and the rebuild for each.
More checklists ship every quarter
Permission sprawl audit, data-quality scorecard, and a pre-deploy review checklist are in the roadmap below.
Quick answer
A cheatsheet is a reference (look up a number). A checklist is an audit (walk top to bottom, mark each item). Use a checklist when you want a structured exercise that produces a list of findings; use a cheatsheet when you just need to recall a specific limit or pattern.
About these checklists
Most Salesforce problems aren't architectural. They're 15 small things that all individually look fine and collectively kill the org. The Flow that worked yesterday hits a governor limit today because data volume tripled. The permission set that looked harmless in May exposes a PII field in August because someone added a new field. The schedule-triggered Flow with no filter scans the whole record base every hour because nobody set the modified-date condition.
These small things are detectable. They're not detectable by reading the docs. They're detectable by walking through a checklist with your laptop open. So we built the checklists. Each one was designed by an engineer who debugged the failure mode in production at least three times.
Technical debt is a daily challenge for many admins. Most of that debt is a few hundred small anti-patterns piled up. A checklist surfaces them in an hour; finding them through bug reports takes years.
If you read a checklist and find zero items in your org, we shipped a bad checklist (or you have a remarkable org). The bar is that any production Salesforce org over 50 users will have at least 3 hits per checklist. If yours doesn't, write it up; we'll feature the case study.
We have rules for what makes the cut. If a checklist doesn't pass all four, it doesn't ship.
No 'review your security settings' generalities. Every item is one specific anti-pattern (e.g., 'DML inside loops') with a Signal you can grep for, a reason it tanks, and a Rebuild that fixes it.
Total run time is bounded (10 to 30 minutes). If a checklist needs three afternoons, it's actually a guide pretending to be a checklist.
Output of running a checklist is a list of items marked Found vs. N/A. The findings list is what gets prioritized into remediation work, not the checklist itself.
Every item came out of an actual production failure we debugged. We don't ship checklists that anticipate hypothetical issues; the bar is 'we've seen this break a real org at least three times.'
Checklists fail when readers treat them as inspiration. They work when readers treat them as a structured exercise with a deliverable.
Three patterns that compound
Don't read the checklist first and 'plan to run it later.' Open Setup, open the checklist, and walk top to bottom in the same sitting. Each item should take under 2 minutes to evaluate.
Print the checklist or copy it into a doc. Marking is the forcing function that turns a read into an audit. Items left unmarked don't count.
Don't try to fix all findings at once. Sort the Found column by blast radius (DML in loops on a high-volume object > hardcoded ID in a low-traffic Decision). Fix the top 3 first; queue the rest.
What not to do
Each new checklist comes from the third or fourth time we run the same audit by hand. The next batch covers the audits we run most often.
Salesforce Permission Audit Checklist
20-item audit of profile + permission set sprawl, with the 5 most common over-privileged patterns.
Salesforce Data Quality Checklist
12-item audit covering duplicate rate, fill rate, validation rule coverage, and silent ETL failures.
Pre-Deploy Sandbox Checklist
The 18-item walk-through every change should pass before promoting from sandbox to production.
Quarterly Org Health Checklist
30-minute audit covering license utilization, storage, integration health, and the top 10 Flow anti-patterns.
If a checklist you'd value isn't here, send it via the contact form. The third or fourth request for the same audit moves it to the front of the queue.
Six adjacent surfaces inside the Clientell library. Each one shows the same Salesforce topic from a different angle. Pick the format that matches the work you're about to do.
Frequently asked
A cheatsheet is a reference (look up a number, recall a pattern). A checklist is a structured audit (walk top to bottom, mark each item, produce a findings list). The same content can be either format depending on the use; we publish the format that matches how the artifact is actually used in practice.
On-page reading is free. PDF download requires a work email. The PDF and the on-page version are byte-identical content; the PDF exists for offline / printable use during an audit.
Each checklist tells you. Quick audits (Flow anti-patterns, permission spot-checks) run monthly. Deeper audits (org health, quarterly business review) run quarterly. Pre-deploy checklists run on every deploy.
That's normal on the first run. Triage by blast radius: high-volume object + governor-limit risk goes to the top of the queue; low-traffic Decision goes to the bottom. Don't try to clear a 50-item findings list in one sprint.
Most checklists target Enterprise Edition by default. Where the rules differ in Professional / Unlimited, the item calls it out inline. Industry-specific checklists (HIPAA, financial services) call out their scope at the top.
Yes. Share the on-page URL for live reference. Print or download the PDF for the audit itself; people mark it up better on paper.
Yes. Send the artifact (the actual list of items, with Signal / Why it tanks / Rebuild for each) plus a 1-paragraph context note about which audit it serves. Quality bar: 10 to 30 distinct items, each runnable in under 2 minutes, all sourced from real production debugging.
Yes, as part of a paid engagement. Book a scoping call via the demo link. We'll usually start with our published checklists, run them against your org, and customize the gaps. Custom checklists become reusable artifacts your team can run quarterly.
Checklists surface the issues. Fixing them by hand is where time still gets burned. Hand the rebuild work to Clientell AI and clear a 50-finding audit in days, not quarters.
14-day free trial · No credit card required