Work & career
Write your ideal job description: must-haves, nice-to-haves, deal-breakers, then test it against real postings
An ideal job description is a one-page description of the job you want, written the way an employer writes a posting: a title, the main tasks, the context, and your criteria sorted into must-haves, nice-to-haves and deal-breakers. Written down, a vague wish becomes something you can test: you score real postings against it, see which criteria the market rarely offers, and adjust. This guide gives the template, the arithmetic of how many must-haves to keep, a worked scoring of five postings, what to do when you cannot fill it in yet, and how to use it in the job you already have, through what researchers call job crafting.
By the Dream Architect team · Published October 22, 2026
The template: write it like a posting
Use the headings that real descriptions use. O*NET, the occupational database published by the US Department of Labor, describes each occupation with sections such as Tasks, Work Context and Wages & Employment Trends; a posting adds the place, the team and the conditions. Borrow that structure, but write it about the job you want.
Fill in the table below in 20 minutes, without editing. The criteria rows are the part you will test, so be specific: “fully remote” and “remote two days a week” are different criteria. For the pay floor, the guide A job that pays the bills works out the salary your budget needs.
| Line | What to write | Question it answers |
|---|---|---|
| Title | The role as postings name it, not a slogan | What would I type into a job board? |
| Purpose | One sentence: what the job exists to do | Why does this job exist? |
| Main tasks | Five tasks, with the share of the week each takes | What would I actually do on a Tuesday? |
| Context | Place, schedule, team size, remote or on site | Where and with whom? |
| Pay floor | The lowest pay that covers your budget | What is the minimum that works? |
| Must-haves | At most three | Without what would I refuse? |
| Nice-to-haves | As many as you like | What would tip a choice between two offers? |
| Deal-breakers | Two or three | What would make me leave within a year? |
Must, nice, no: why at most three must-haves
Each must-have shrinks the number of postings that qualify, and they multiply. A simplified calculation of ours, assuming each criterion is met by half of the postings in your field and that criteria are independent of each other: one must-have leaves 1 posting in 2, three leave 1 in 8 (0.5 × 0.5 × 0.5 = 12.5%), five leave 1 in 32 (about 3%). Real criteria are rarely independent, and some are met by far fewer than half, so treat this as an order of magnitude, not a forecast.
Hence the rule: keep at most three must-haves, move everything else to nice-to-have, and keep deal-breakers for what would truly make you leave. If a criterion feels important but you would accept a good offer without it, it is a nice-to-have.
Test it against real postings
Read ten real postings for the title you wrote and score each one: must-haves met (out of three), nice-to-haves met, any deal-breaker. A posting that meets every must-have and no deal-breaker is worth applying to; a criterion that no posting meets is a signal about the market, not about you.
The table is an illustrative example, not a real person or real postings: someone aiming for a data analyst role, with three must-haves (remote at least two days a week, a team of at least three analysts, a pay floor of £38,000 a year) and one deal-breaker (regular weekend on-call).
| Posting | Must-haves met | Nice-to-haves met | Deal-breaker | Decision |
|---|---|---|---|---|
| A | 3 of 3 | 2 of 4 | None | Apply this week |
| B | 3 of 3 | 0 of 4 | None | Apply |
| C | 2 of 3 (fully on site) | 3 of 4 | None | Ask about remote days before applying |
| D | 3 of 3 | 1 of 4 | Weekend on-call | Skip |
| E | 1 of 3 | 1 of 4 | None | Skip |
If you cannot fill it in yet: three ways to get clearer
Clarity usually comes from information, not from thinking harder. Three sources, each a few hours at most:
- Your last two weeks. List the tasks you did and mark each one: gave energy, neutral, drained. The tasks that gave energy are candidates for the Main tasks line.
- The official description of the occupation. Look up the title in O*NET and read its Tasks list; tick the tasks you would like to do most days and strike the ones you would not.
- Two people who do the job. Ask each the same question: “What does an ordinary Tuesday look like?” Their answers will fill the Context line better than any posting.
Job crafting: use the description in the job you have
You may not need a new employer to close part of the gap. In 2001, Amy Wrzesniewski and Jane Dutton described job crafting: the changes people make on their own initiative to the tasks of their job, to their relationships at work and to the way they think about the job. In a 2010 Harvard Business Review article with Justin Berg, they turned this into an exercise that, in their words, “prompts you to visualize the job, map its elements, and reorganize them to better suit you.”
The evidence on organised crafting programmes is mixed. In a randomised controlled trial with 281 employees in Japan, published in 2020, two 120-minute group sessions a month apart had no significant effect on work engagement at three or six months; a small effect at three months appeared only among people who crafted little at the start, in an analysis the trial had not planned. Crafting is worth trying because it costs little, not because a programme reliably works.
Compare your ideal description with your current job and pick one gap. The table shows one crafting move per kind of change.
| Kind of crafting | Gap you notice | One move this month |
|---|---|---|
| Tasks | You want more analysis, you spend Fridays on reports | Offer to automate one report and use the hour for an analysis |
| Relationships | You want to learn from senior people | Ask one senior colleague for a 20-minute review of your work |
| Perceptions | The work feels pointless | Write who uses your output and what it changes for them |
Turn the description into a nightly scene
The Main tasks line is the raw material for a process scene: one ordinary moment of the job you described, not the day you sign. In Dream Architect, the Work & purpose area offers 4 pre-written scenes among the app’s 84 in 21 life areas (counted in the app’s content on 22 October 2026); for this, use “Write my own moment” and write two present-tense sentences from your description: “Tuesday, 10 a.m., I open the dataset, my two colleagues are on the call, and I explain the first chart.”
The guide How to get your dream job, step by step turns the description into a weekly application rhythm, and Career scripting shows how to write a full future workday from it.
What the description can and cannot do
It makes your search specific, gives you a way to compare offers, and shows which criteria the market rarely meets. It does not create postings that do not exist, and writing it does not make an employer offer it. Update it after every ten postings you read: the market is part of the answer.
Questions people ask next
What should an ideal job description include?
A title as postings name it, a one-sentence purpose, five main tasks with their share of the week, the context (place, schedule, team), your pay floor, at most three must-haves, your nice-to-haves and two or three deal-breakers.
How many must-haves should I keep?
At most three. Each must-have cuts the postings that qualify, and the cuts multiply: if each criterion were met by half the postings, three would leave about 1 in 8 and five about 1 in 32.
What if no posting matches my description?
Find which criterion no posting meets. Either the market rarely offers it, and you move it to nice-to-have, or you are searching under the wrong title; read the official description of nearby occupations and ask two people in the field.
What is job crafting?
Job crafting is the set of changes people make on their own initiative to their tasks, their relationships at work and the way they see their job. The term comes from Amy Wrzesniewski and Jane Dutton (2001).
Does job crafting work?
Evidence on programmes is mixed: in a 2020 randomised trial with 281 employees in Japan, two group sessions had no significant effect on work engagement overall; a small effect among people who crafted little at the start came from an analysis the trial had not planned. Small personal changes cost little, so they are worth testing one at a time.
How do I get career clarity if I do not know what I want?
Gather information instead of thinking harder: mark which of your recent tasks gave or drained energy, read the official task list of a few occupations, and ask two people what an ordinary Tuesday in their job looks like.
Try it tonight
- Write the title of the job you want the way a posting would name it, and one sentence on what it exists to do.
- List your criteria, then keep at most three must-haves and move the rest to nice-to-have.
- Picture one ordinary Tuesday moment of that job for two minutes, from your list of main tasks.
How we made this guide
Written by the Dream Architect team with AI assistance, then checked line by line against the app’s actual screens and the sources below. Third-party facts link to their source. Figures about the app were counted from its content on October 22, 2026.
- Wrzesniewski & Dutton (2001), Crafting a job: revisioning employees as active crafters of their work, Academy of Management Review 26(2), 179–201 (accessed October 22, 2026)
- Wrzesniewski, Berg & Dutton (June 2010), Managing yourself: turn the job you have into the job you want, Harvard Business Review (accessed October 22, 2026)
- Sakuraya, Shimazu, Imamura & Kawakami (2020), Effects of a job crafting intervention program on work engagement among Japanese employees: a randomized controlled trial, Frontiers in Psychology 11, 235 (PubMed Central) (accessed October 22, 2026)
- O*NET OnLine (US Department of Labor, Employment and Training Administration), occupation summary with Tasks, Work Context and Wages & Employment Trends, example: Data Scientists 15-2051.00 (accessed October 22, 2026)
This guide describes a personal reflection habit. It is not medical or psychological advice and does not promise any particular result.