On a slow Thursday afternoon, the operations manager of a small charter brokerage decided to do something she had put off for two years. She closed the quote window, brewed a strong tea and opened a blank page titled “Who can touch what?” The result was a one-afternoon audit of her team’s digital accounts, and it is the clearest introduction to IdentityOps that most charter desks will ever need. The character is invented, yet every step below is real work you can repeat. Print it, tick the boxes, and note the time each step takes.
Before you start the IdentityOps audit
Gather three things. The current staff roster, the list of tools the desk pays for, and a quiet hour. The roster is your reference for who should exist; the tool list tells you where accounts live. Remember what the Ops Atlas says about scope: the work centres on the existence of each identity and the proof behind it, while rules about what each one may do belong elsewhere. Keeping the audit narrow is what lets it fit into an afternoon.
Checklist part one: the people
- Export user lists. From the quoting tool, the payment portal, the shared drive and the messaging service, export the list of accounts. If a tool cannot export one, take screenshots.
- Match against the roster. Mark every account without a current employee. She found six, including a flight-planning login for a contractor whose project had ended in the spring.
- Look for shared names. Anything called “ops”, “desk” or “sales” is probably used by several hands. Plan to replace each with named seats.
- Check the last-used date. The Atlas suggests treating accounts idle for roughly three to six months as candidates for disabling, excluding people on formal leave and emergency logins.
Checklist part two: the doors that open to strangers
- Review second factors. Does everyone have one? Which type? A text message code is weaker than a security key. The entry names hardware keys and certificate cards as the only methods that resist phishing, and suggests starting with whoever can change bank details or release a trip.
- Test recovery. Ask yourself how a colleague with a lost phone would get back in. If the honest answer is “they ring me and I reset it”, you have found your biggest gap. Write a three-line script that uses a second method first, then a manager’s confirmation, never a voice on the phone alone.
- Look at connected apps. In your payment and mail tools, open the list of authorised third-party apps. Strike what you do not recognise, and give owners to what remains.
The story behind step six
Why the fuss about recovery and connected apps? Because the public record is full of incidents in which a routine, not a code, failed. The Atlas summarises several. In one, an intruder guessed passwords at scale against a legacy test environment lacking a second factor and then used a forgotten test application to reach mailboxes. In another, access to a data-warehouse customer base was gained through credentials that had not been changed for as long as four years and sat on accounts without a second factor. A third group relied on talking help desks into resetting passwords and swapping SIM cards. A charter desk trades in high-value, time-critical requests, which makes it a believable target for exactly that kind of social approach.
Checklist part three: the leavers
- Pick the last person who left. Trace them through every tool. How long did full removal take?
- Cancel open sessions. Turning off an account may leave devices signed in; make it a standard step.
- Do the arithmetic honestly. If four of the last five leavers vanished from every tool, report 80 percent, not the friendly figure for the main directory alone.
- Write the target. An example from the Atlas: central login disabled within four hours, all applications within a day.
Checklist part four: the robots
- List automated accounts. Price feeds, calendar syncs, and anything a developer once built to export trips.
- Name an owner and an expiry. If nobody wants the job, that account is an orphan.
- Shorten the life. Replace a never-changing key with one that is rotated on a schedule. To retire one, disable it, wait, and only then delete.
Checklist part five: keep it alive
- Daily: let a script, or a person with a coffee, compare the roster with the accounts and flag mismatches.
- Weekly: pick two recent account recoveries and confirm each followed the script.
- Monthly: list which automated credentials expire in the next sixty days.
- Quarterly: stage a pretend departure and review which outside sign-in services you still trust.
This rhythm comes from the Atlas, and its attraction is how little each step demands. None of the four tasks takes longer than a flight briefing, yet together they stop the slow drift that turns a tidy register into a guess. Small desks can simply merge the weekly and monthly items into one sitting.
What she did with the page
By five o’clock the page held twenty-two crossings-out and three phone calls she needed to make. She put a quarterly reminder in her calendar titled “Who can touch what?” and gave the page a home next to the safety procedures. Pricing and routing are the glamorous work of a charter desk, as described in the article on the factors that push a private jet booking up or down and in the explainer on how repositioning flights are sold. Yet every quote depends on logins behaving.
If you would like the full reasoning, including maturity levels and a thirty-sixty-ninety plan, the Ops Atlas entry for IdentityOps is a good place to read next. Tonight, though, finish the first checklist and put the kettle on.

