Proxy work
busywork · productive procrastination
Activity that resembles progress and substitutes for it. Proxy work is comfortable because it is legible and low risk, and it reliably crowds out the one uncomfortable task that would actually move the number.
In practice
A full day cleaning the pipeline data. Real work, needed doing, and not one prospect was contacted. The pipeline is tidier and no closer to producing revenue.
The common mistake
Judging a day by how much was completed. Proxy work always completes, which is part of its appeal, and completion is not the same measure as movement.
There is a category of task that is real, useful, defensible out loud, and chosen precisely because doing it means not doing the other thing. It is not laziness, which is why it survives every attempt to fix it with discipline.
Why it is not procrastination
Piers Steel's meta-analysis (2007) found procrastination predicted principally by task aversiveness, impulsiveness and delay to reward — the procrastinator does something obviously unproductive and knows it. Proxy work differs on the one variable that matters: it produces a completed thing. The pipeline really is cleaner. John Perry (1996) noticed the same structure from the other side and called it structured procrastination — the avoidance of one task reliably generates output on several others, which is why the habit is self-reinforcing rather than self-correcting.
Why it survives review
The reason organisations do not catch it is that they mostly cannot. Erin Reid (2015) studied a global consultancy where an eighty-hour norm was enforced, and found that employees who quietly worked around fifty hours while passing as fully committed received performance ratings indistinguishable from those who genuinely worked eightyReid, E. (2015). 'Embracing, Passing, Revealing, and the Ideal Worker Image.' Organization Science 26(4). The finding is not that the passers were better at their jobs, but that the organisation was rating a projected image of effort rather than either effort or output — and could not tell the difference.. The organisation was assessing a projected image of effort and could not distinguish it from the thing itself. If a review process cannot separate the appearance of work from work, it will select for the appearance, which is Goodhart's law operating on the individual rather than the metric.
Leslie Perlow (1999) documented the environmental half. Engineers in her study lost their productive hours to an interaction pattern that made concentrated work nearly impossible, and the visible, interruptible, responsive tasks expanded to fill what was left — Parkinson's observation (1955) with a mechanism attached. Proxy work is partly a choice and substantially a structure.
The objections
The category is unfalsifiable after the fact. Any activity that did not directly produce revenue can be relabelled proxy work in hindsight, and most genuinely necessary work — maintenance, documentation, relationship-building, thinking — is indirect by nature. Used carelessly the concept becomes a instrument for bullying yourself out of everything except selling, which is a good way to run a business into a different wall.
David Graeber's Bullshit Jobs (2018) made the strongest popular version of the argument and is worth reading with the caveat that its empirical base is thin — a YouGov poll and self-selected testimony, which is close to the worst available evidence about whether work is useful, since the people best placed to judge are the least disinterested.
What it rules out
It rules out completion as the measure of a day. Proxy work always completes, which is the whole of its appeal. It rules out a full calendar as evidence of progress. And it rules out treating the problem as motivational when the review process cannot distinguish output from visible effort, because then it is structural and no amount of discipline touches it.
It does not rule out indirect work. The usable test is prospective rather than retrospective: decide before the day what would count as movement, then see what got done. Deciding afterwards makes the category unfalsifiable and useless.
Sources
Graeber, D. (2018). Bullshit Jobs: A Theory. Simon & Schuster. · Parkinson, C. N. (1955). 'Parkinson's Law.' The Economist, 19 Nov. · Perlow, L. (1999). 'The Time Famine: Toward a Sociology of Work Time.' Administrative Science Quarterly 44(1). · Perry, J. (1996). 'Structured Procrastination.' Chronicle of Higher Education. · Reid, E. (2015). 'Embracing, Passing, Revealing, and the Ideal Worker Image.' Organization Science 26(4). · Steel, P. (2007). 'The Nature of Procrastination.' Psychological Bulletin 133(1).
Concept web
Open the full webQuestions
What is proxy work?
Activity that resembles progress and substitutes for it — real, useful and defensible, but chosen because doing it means not doing the harder thing that would move the outcome.
How is proxy work different from procrastination?
Procrastination produces nothing and the procrastinator knows it; Steel's meta-analysis (2007) ties it to task aversiveness and impulsiveness. Proxy work produces a completed thing, which is why it survives self-review and why discipline does not fix it.
Why don't organisations catch proxy work?
Because most review processes cannot distinguish work from the appearance of work. Reid (2015) found employees who worked around fifty hours while passing as eighty received indistinguishable performance ratings from those who genuinely did.
How do you tell proxy work from necessary indirect work?
Decide in advance what would count as movement, then check what was done. Judged afterwards the category is unfalsifiable — almost any non-revenue task can be relabelled proxy work in hindsight, including the maintenance the business needs.