Every IT team has The Spreadsheet. Forty-seven tabs. Color-coded by department. Columns for every system, every group, every permission a new hire might need. Notes in the margins: “Check with Sarah first” and “Only for senior engineers” and “This might be outdated.” Last modified: eight months ago. By someone who left the company six months ago. This spreadsheet serves as the source of truth for what access new employees should get. It’s also wrong. Everyone knows it’s wrong. But fixing it is a project nobody has time for, and besides, it’s “close enough” most of the time. That holds until the day it doesn’t.Documentation Index
Fetch the complete documentation index at: https://docs.provisionr.io/llms.txt
Use this file to discover all available pages before exploring further.
The Baseline Entitlement Problem
In identity management, you may see the term “birthright entitlements” that refer to the access that all employees should get as an employee of a company of a member of a department. In other words, being a “citizen” of the organization entitles you to certain baseline access. We use the term “baseline entitlements” instead, because “birthright” has connotations we want to avoid, and because what matters most is the baseline access associated with a role. You may also hear this as “role-based access,” but that term is overloaded in identity management circles. Each of your organization profile attributes can be used to provide you baseline access to various systems. This is also referred to as Attribute-Based Access Control (ABAC). If you’re in the Marketing department, you get the marketing Google Groups, the marketing Slack channels, the marketing shared drives. That’s your baseline as a member of the Marketing department. The problem is how organizations track and provision baseline entitlements: manually curated checklists that are always incomplete and always outdated. Here’s why this breaks down:Systems evolve faster than documentation
A team adopts a new tool and adds it to their mental checklist. Does anyone update The Spreadsheet? Sometimes. Eventually. Perhaps never.
Organizations restructure faster than processes can follow
Sales splits into two regional teams. Product creates a new sub-group. Engineering adds a platform team. Each reorganization should trigger a checklist review. Each reorganization doesn’t.
Institutional knowledge walks out the door
The person who knew why row 47 existed left last year. The person who maintained the “Engineering” tab got promoted and doesn’t have time anymore. Knowledge decays.
The Hidden Cost of “Close Enough”
“Close enough” carries real costs that don’t appear on any dashboard:Lost productivity compounds daily
A new engineer can’t access the code repository on day one. They can’t complete onboarding tasks. They can’t contribute. A senior engineer billing at $150/hour who’s blocked for two days waiting for access represents $2,400 of lost productivity. Multiply that by every new hire, every year.
IT overhead scales with headcount
Every missing checklist item becomes a ticket. Every ticket takes time to process. If the onboarding checklist is 80% accurate, dozens of tickets per new hire cover the missing 20%. That’s not exceptional workload—that’s predictable, preventable workload.
First impressions shape retention
New hires form opinions about a company in their first week. When that first week is spent fighting for access to basic tools, the impression isn’t “this company has its act together.” It’s “I wonder what else is broken here.”
Security risk cuts both ways
Missing access frustrates employees. Incorrect access—permissions granted that shouldn’t be—creates security exposure. When checklists are poorly maintained, both problems occur.
Why “Just Update the Checklist” Fails
The obvious answer to checklist problems is “keep the checklist updated.” Create a process. Assign an owner. Review quarterly. This works for about six weeks. Then someone’s on vacation when a system change happens. Then a reorg creates three new teams and nobody remembers to update the tabs. Then the quarterly review gets pushed because everyone’s busy with end-of-quarter work. Then someone updates a row but doesn’t tell anyone. Then someone else reverts it because they didn’t know it changed. Checklists fail because they require continuous human effort to maintain state that exists elsewhere in organizational systems. The information about what teams exist, what systems those teams use, and what access those systems require—that information already exists. It’s in the HRIS, the IdP, the application configurations. A checklist is a manual copy of information that should be authoritative somewhere else. Manual copies always drift from the source.From Checklists to Policies
What if, instead of maintaining a checklist of “what to provision,” organizations defined policies for “who should have access”? The difference is subtle but powerful: Checklist thinking: “When someone joins Marketing, add them to these 15 groups.” Policy thinking: “The Marketing Google Group should contain everyone in the Marketing department.” With checklist thinking, actions are scripted for humans to take. Every new hire triggers a manual process. Every change to what Marketing needs requires updating the script. With policy thinking, desired state is defined. The system continuously ensures the policy is true. When someone joins Marketing, they’re automatically in the Marketing Google Group because the policy says Marketing members belong there. This is what a Ruleset does. A Ruleset is a membership criteria policy—a definition of who should have access based on their attributes, not a list of actions to take. Here’s how it works conceptually:Handling Complexity with Multiple Rules
“But access isn’t that simple,” IT teams often think. “It’s not just department. It’s department AND role AND sometimes location AND occasionally exceptions exist.” Rulesets handle this through multiple rules and conditions: Multiple conditions (AND logic):The Checklist That Updates Itself
Here’s the mental model shift: stop thinking of onboarding as “run through a checklist of actions” and start thinking of it as “ensure the user’s attributes trigger the right policies.” When onboarding is policy-based:- HR enters the new hire in the HRIS with their department, team, title, manager, and location
- That data syncs to the directory (Okta, Google Workspace, etc.)
- The policy engine evaluates all Rulesets against the new user’s attributes
- The user automatically matches every Ruleset their attributes qualify them for
- Membership syncs to downstream systems (Google Groups, Slack, GitLab, etc.)
- By the time they sign in on day one, access is already provisioned
Migration: From Spreadsheet to Rulesets
Existing checklists don’t need to be thrown away. They can be migrated:Audit the checklist
For each row, identify: What group or resource? What criteria determines access? Is this still accurate?
Group by pattern
Most checklist items fall into patterns: “everyone in department X,” “everyone with title Y,” “everyone reporting to manager Z.” Group similar items together.
Create Rulesets for patterns
Instead of 47 checklist items that all say “add to Marketing groups,” create one Ruleset for Marketing team membership with appropriate conditions.
Handle exceptions explicitly
Some checklist items are genuine exceptions—specific people who need access outside normal patterns. Add these as named user rules.
Validate before activating
Use staged rules to preview who would get access. Compare against current group membership and the existing checklist. Identify discrepancies before going live.
What Happens When the Checklist Is Gone
Teams that replace checklists with policies report consistent outcomes:Onboarding time drops
When access is pre-provisioned, new hires are productive on day one. No waiting for IT to work through a checklist. No discovering missing access throughout the first week.
Access request volume drops
When baseline entitlements are automatic, the “I just joined the team and can’t access anything” requests disappear. IT can focus on actual exceptions.
Audit confidence increases
“Show me who has access to X and why” becomes a database query instead of a forensic investigation across spreadsheets and ticket history.
In the next article, we’ll tackle an even harder problem: what happens when employees don’t leave—they just change roles. The “mover” problem creates compliance risks that most organizations don’t even realize they have.