Job crafting

Job crafting is the finding that two people with the same job description do not have the same job, and that the difference is partly authored.

Amy Wrzesniewski and Jane Dutton named it in 2001 after noticing something in a hospital. Cleaners with identical duties described their work in incompatible ways. Some described cleaning. Others described being part of the care — noticing which patients had no visitors, timing their rounds so a long-stay patient had company in the afternoon, rearranging pictures in rooms where people could not move. Nothing in the job description had changed. The job had.

The idea before this was that motivation is engineered from above: Hackman and Oldham's job characteristics model, still the standard, treats variety, significance and autonomy as things a designer builds into a role. Job crafting is the claim that the occupant is also a designer, usually without noticing, and often better positioned than the one with the org chart.

Wrzesniewski and Dutton's original account has three levers, and they are worth keeping separate because they cost different amounts.

Task crafting. Changing what you actually do — taking on the part of the work you are good at, trading away the part someone else does faster, adding something that is not in the role because nobody owns it. This is the most visible and the one most likely to need permission.

Relational crafting. Changing who you do it with and how. Which colleagues you build a working relationship with, which clients you take, whether the handover is an email or a conversation. This is nearly always available and almost never discussed, which makes it the cheapest lever most people are not pulling.

Cognitive crafting. Changing how you frame the whole. The cleaner who is part of the care rather than the cleaning; the support engineer who is in the business of someone's worst morning rather than ticket throughput. This one is free, and it is also the one most easily counterfeited — a story told about work that has not changed at all is not crafting, it is coping.

A fourth lever entered later through the job demands-resources tradition. Tims, Bakker and Derks built a scale around increasing resources and challenges, and decreasing hindering demands — which is the same idea in the vocabulary of load, and the version most of the quantitative research uses.

Rudolph and colleagues meta-analysed job crafting in 2017. The picture is real but should be stated with its limits attached.

Crafting — particularly the approach-oriented kinds, increasing structural resources and challenges — relates to engagement, job satisfaction and performance, and to lower burnout. Reducing hindering demands behaves differently and sometimes worse, which is the finding people skip: quietly removing the parts of the job you dislike is not the same intervention as adding the parts you want, and it does not show the same benefits.

Two honest caveats.

Most of the evidence is cross-sectional. People who craft more report better outcomes; the arrow is probably bidirectional, since it is easier to redesign a job you have energy for. The longitudinal and intervention studies are fewer and their effects are smaller than the correlations suggest.

Crafting needs slack. The same meta-analysis found crafting related to job autonomy. A role with no discretion, no spare hour and a manager who counts keystrokes is not craftable, and telling someone in one to craft is close to an insult. The proactive personality correlation points the same way: this is easier for people already inclined to act, which means it under-describes the situation of everyone else.

This entry would be dishonest without the next paragraph.

Job crafting is the most convenient concept in organisational psychology for an employer, because it relocates a design problem into the employee. A role that is badly scoped, under-resourced or meaningless is a management failure; "have you tried crafting it?" converts that failure into a personal development opportunity, at no cost. The research does not claim this, and Wrzesniewski's own work is careful about the conditions crafting requires. Trainings built on it often are not.

The test is simple. Crafting works on the shape of a job and not on its terms. It can change which tasks you hold, who you do them with and what they mean to you. It cannot fix pay, a hostile manager, a headcount that is half what the work needs, or a company that is going to miss payroll. If the problem is in the second list, crafting is a way of spending your own energy to subsidise something that is not yours to fix, and the burnout entry describes where that ends.

Two live questions keep the distinction honest: if this worked perfectly, would I want to still be here in two years? and is the thing I am crafting around a task, or a person?

Start from the fifteen percent. Almost every job contains a slice that is genuinely at your discretion — the order you do things in, which of two projects you volunteer for, whether you write the doc nobody asked for. Crafting is what you do with that slice, and it works better as an accumulation of small moves than as a conversation with your manager about your role.

Some specifics that tend to hold:

• Trade rather than drop. "I'll take your client calls if you take my reporting" is a proposal. "I would like to stop doing reporting" is a request, and a worse one.
• Take the adjacent problem. The fastest route to a changed job is owning something that nobody currently owns, because there is no incumbent to displace and no permission to seek.
• Craft the people first. Relational moves need the least sign-off and change the daily experience the most. One good working relationship converts a series of transactions into a set of conversations.
• Name the meaning to yourself, once, in specific terms. Not "I help people" but the actual person your work reached last month. Cognitive crafting fails when it stays at slogan altitude.

If you do not know which direction to craft in, the missing input is usually values rather than effort. The Career Values test reads what you are actually optimising for, and the Career Interest Match reads the kind of work you are drawn to — both are more useful before a crafting attempt than after one. And if the answer is that no available shape of this job is one you want, that is a legitimate result, and sunk cost in careers is the next entry rather than this one.