How many can you afford to lose?

How many people could your project lose — to a bus, a better offer, a long holiday — before some part of it has nobody left who knows it? Map who knows what, and this finds the answer: the fewest departures that leave an area orphaned, plus which areas and which people are the single points of failure.

A
B
C
D
Area × personRate each person's depth in each area: — none, 1 learning, 2 knows it, 3 owns it. Someone counts as able to keep an area alive on their own at depth 2 or more; “learning” is bench strength but can't hold it alone yet. The bus factor is the fewest such knowers any one area has.
ABCDdepthHow many people could keep this area alive on their own (depth ≥ 2). 0 is a gap nobody owns; 1 is a single point of failure; 2+ has a backup.
2
1+1
1+1
2+1
2
1+1
1bus factor

Your bus factor is 1. Lose the wrong person and 3 areas (Billing, Infra / deploy, Data model) loses its only owner. The most exposed is Ana — gone, and 2 areas go dark at once.

owners per area
Billing
1 owner +1 learning
Infra / deploy
1 owner +1 learning
Data model
1 owner +1 learning
Auth
2 owners
Search
2 owners +1 learning
Mobile app
2 owners
areas orphaned if they leave
AAna
2 areas dark · owns 3
DDev
1 area dark · owns 2

These people are the only ones at working depth in some area — the quickest way to raise your bus factor is to get a second person to depth 2 in the areas above.

bus factor = the fewest people whose loss leaves an area with no owner · depth is your honest read, so the number is only as good as it