The Master Key That Was Never Returned
A 40-employee company calls you after a manager is terminated. The manager held a master key that opened all offices, the server room, and the exterior doors. HR believes the key was returned, but no one can confirm it, and the key was never signed for on a key-control log - the company has no key register at all. They want you to 'rekey the master' so the old key stops working, cheaply, over a weekend, without disrupting staff on Monday. The building has 40 office cylinders, 6 exterior doors, and a server room on a separate high-security cylinder. You have not seen the existing pinning.
1. What does 'rekey the master so the old key stops working' actually require here, and why is it more than a single cylinder change?
Read the worked answer
A master key opens every cylinder in its domain, so to make that key stop working you must change every cylinder the master operates - all 40 office cylinders and the 6 exterior doors at minimum, and you must address the server room's separate high-security cylinder on its own terms. There is no single-cylinder fix: changing one lock leaves the master valid on the other 45, which is the opposite of what the company needs.
It also requires a new hierarchy, not just new cuts: a new master and a new set of change keys, built on a documented progression, so the old master matches nothing. And it requires a key register going forward, because the reason they are in this position is that no one can say whether the key came back. 'Rekey the master' is shorthand for 'rebuild the keyed system and its control,' and scoping it as one cylinder change would be a serious misjudgement.
2. The company wants it done cheaply over one weekend without Monday disruption. What is a realistic plan, and where do you push back?
Read the worked answer
Realistic plan: pre-build and pre-cut everything off-site before the weekend - pin kits for all 46 cylinders, the new master and change keys, and the server-room cylinder handled per its high-security requirements. On site, work in a sequenced sweep (exterior doors early, offices in blocks, server room with the company's IT witness), test each cylinder against its change key, the new master, and confirm the old master fails, then hand over a key register with every key signed out.
Where you push back: (1) 'cheaply' cannot mean reusing the old progression or skipping cylinders - a partial rekey leaves the terminated manager's key live somewhere, which defeats the purpose; (2) a weekend may not be enough for 46 cylinders plus the high-security server room if the existing pinning is undocumented and has to be mapped first, so you scope the site survey before committing to a date; (3) the server room may require a restricted or high-security solution that costs more than the office cylinders, and you say so up front rather than substituting a cheaper cylinder that undermines the point. You give them the honest number and timeline, not the one that wins the job.
3. What do you recommend about key control going forward, and why is it part of the job rather than an upsell?
Read the worked answer
You recommend and set up a key register: every key numbered and recorded to a holder, signed for on issue and on return, with masters and the server-room key tracked separately and limited to named individuals. You recommend the new system be a restricted keyway so future duplication requires authorization, and you recommend a documented pinning chart held by the company (and by you) so the next rekey does not start from zero.
Why it is part of the job: the immediate problem - an unreturned master no one can confirm - was caused by the absence of key control, not by the lock. If you rekey and hand back 46 cylinders' worth of keys with no register, you have reproduced the exact conditions that produced this call, and the next termination will produce the same emergency. Delivering a rekey without delivering control is delivering half the job; the register and the restricted keyway are what make the rekey hold.