Nobody in your organisation wants a weekly status meeting. Nobody wants a project tracker, a Slack channel, or a quarterly planning template either. These things were all hired — brought in at some specific moment to get something specific done — and the moment has usually been forgotten while the employment continues.
Clayton Christensen, the Harvard strategist best known for his theory of disruptive innovation, spent his later career on a deceptively simple reframing of demand. People do not want products as such, he argued; a job arises in their life, in a particular circumstance, and they look around for something to get it done. His research team's famous example involved a fast-food chain's morning milkshakes: sales data described the buyers, but only watching the circumstance revealed the employment — commuters with a long, boring drive were engaging a breakfast that could be managed one-handed and would last the journey. The competition was not other desserts. It was bagels and bananas, and the milkshake was winning on grip and duration.
The unit of analysis, in other words, is not the product and not even the customer. It is the circumstance and the job arising in it. Understand those, and both design and competition snap into focus.
Run the employment review inward
Turn the question on your own operations and it becomes an audit with teeth. For every recurring meeting, tool, report and process, ask three things. What job was this originally hired to do — what was going wrong, for whom, that made someone reach for it? Is that job still arising? And if it is, is this still the best-qualified candidate?
The answers sort everything into four honest piles. Some employments are sound: the job persists and the hire still fits. Some jobs have vanished — the meeting hired to coordinate two teams that merged a year ago — and the position should be closed, not out of austerity but because an employee with no job invents one, usually surveillance or theatre. Some jobs persist but the hire has been outgrown: the tracker adopted at eight people, now failing at eighty, kept because switching feels like an accusation against whoever chose it. And some jobs were never real — the process hired to prevent a disaster that happened once, five years ago, to a different team.
The circumstance test also explains why imported best practices disappoint. Another company's ritual was hired for a job arising in their circumstance; copied into yours, it is an employee who showed up for work that does not exist here. The honest sequence runs backwards from every playbook: name the job first, then interview candidates — and notice that doing nothing is always one of them.
Watch out for
The framework's weakness is that jobs are inferred, not observed, and an inference can be self-serving. The meeting's owner will discover that its job is alignment, a job conveniently impossible to declare finished. Christensen's own discipline was to anchor in the circumstance — what breaks, concretely, in the week this thing disappears? Run the experiment where the stakes allow it: suspend the candidate for a month and watch what actually goes wrong. Silence is a verdict.
His theory also earned real criticism on its home ground — historians pushed back hard on some of the disruption case studies, and he engaged the critics publicly. The jobs lens inherits that lesson: it is a way of asking sharper questions, not a machine that outputs correct answers.
Answer this next
Pick the recurring meeting you most dread. What job was it hired for, in one sentence — and if nobody can produce the sentence, what is it still doing on the payroll?
