Methodology
Every number on this site is an estimate. This page explains what each one means, how we got it, and where it's least reliable. That way you can decide how much weight to give it instead of taking our word for anything.
The short version
Figures are researched ranges for part-time effort in the United States. They come from platform rate data, industry salary surveys, marketplace listings, and practitioner reports. They are not predictions. Most people land at or below the bottom of any range, and a meaningful number earn nothing at all.
How an entry is structured
The index is built in two layers. A line of work is a real, recognizable way people earn money: freelance writing, dog boarding, thrift reselling. An entry is a specific specialization within one of those. We split it that way because the specialization is what actually determines the pay.
"Freelance writing" is not one thing. SEO blog writing, grant writing, and regulatory medical writing have different buyers, different pay, and different skill requirements. Lumping them together would hide all of that, so we keep them separate.
There are currently 1,118 entries across 72 lines of work and 30 categories. That is a smaller number than some sites advertise. We would rather have a thousand entries you can act on than ten thousand generated by pairing every verb with every noun.
What each figure means
| Field | Precise meaning |
|---|---|
| Startup cost | What it realistically costs to get to your first paying customer, in US dollars. Includes equipment, licensing, certification and platform fees. Excludes your time. The low end assumes you buy used and own nothing already; the high end assumes a properly equipped start. |
| Per month | Gross monthly revenue at the stated weekly hours, before tax and before business expenses. Where a range starts at $0, a material share of people doing this earn nothing. That's real data, not a formatting error. |
| Hours per week | The part-time commitment the earnings figure assumes. Includes unpaid work: sourcing, quoting, admin, travel, listing, invoicing. |
| Time to first $ | Typical elapsed days from starting to receiving your first payment, for someone working at it consistently. Includes licensing and approval waits, which are often the bottleneck. |
| Difficulty | A 1–5 rating of skill and knowledge required. 1 means anyone can start today. 3 means a few months of deliberate learning. 5 means professional-grade expertise, a degree, or years in the trade. |
| Remote / In person / Hybrid | Where the work physically happens. Hybrid means a meaningful part of it can be done either way. |
| Active / Semi-passive / Passive | Whether income stops when you do. Passive means an asset keeps earning without ongoing work, once you've done the often substantial work of building it. Semi-passive means it keeps earning with maintenance. |
Where the numbers come from
Figures are compiled from four kinds of source, in roughly this order of weight:
- Published platform data: posted rate ranges, fee schedules, and payout terms from the marketplaces where the work is actually bought and sold.
- Live marketplace listings: what comparable work is being offered and completed for right now, sampled across several markets rather than one.
- Occupational wage data: government and industry salary surveys for the equivalent full-time role. We adjust these down for part-time scale and up for the premium independent providers charge over employees.
- Practitioner reports: what people doing the work say they earn, in trade forums, professional associations, and public income reports. This is the least reliable source. We use it to sanity-check the others, never on its own.
Sources disagree all the time. When they do, we take the more conservative figure. If a source is obviously promotional (say, a course seller quoting the earnings of their own students), we throw it out.
Ranges are set at the line-of-work level
Research is done per line of work, and every entry within it inherits that range unless there is a specific, known reason it differs. Organic chemistry tutoring is set higher than general math tutoring because that difference is well documented. Where no such difference is documented, entries share the range rather than being given invented precision. Two entries showing the same numbers usually means we found no evidence they differ, not that we did not look.
The ease-of-entry score
The default sort is Easiest to start. It ranks entries on a published 0–100 score built from the three things that stop people starting: skill required, money required, and how long you wait to be paid. Nothing else feeds into it. Not our opinion, and not any commercial relationship.
difficultyPenalty = (difficulty - 1) / 4 × 35
costPenalty = log10(1 + startupCostLow)
/ log10(1 + 20000), capped at 1 × 35
waitPenalty = log10(1 + daysToFirstPayment)
/ log10(1 + 365), capped at 1 × 30
easeScore = 100 - (difficultyPenalty + costPenalty + waitPenalty)Costs and waits use a logarithmic scale. To someone starting out, the difference between $0 and $500 matters far more than the difference between $10,000 and $10,500. The weightings (35 / 35 / 30) are a judgment call. We state them here so you can disagree with them. Every entry's score is shown on its detail panel.
This score deliberately says nothing about how much you can earn. Something can be trivially easy to start and still pay very little, which is true of the entry currently ranked first. Sort by Highest earning and the list inverts almost completely. That gap is real, and we would rather show it than hide it.
What the flags mean
Warning flags show up as plain-language notes inside each entry's detail panel, not as badges on the list.
| Flag | What it tells you |
|---|---|
| License required | You need a license, certification, or background check before you can legally be paid. |
| Car needed | Not workable without reliable transport, and often not without commercial-use insurance. |
| Big upfront cost | A realistic start costs more than $1,000. |
| Check local rules | There are real legal, tax, licensing, or insurance questions here. Look up your local rules before you start. |
| Crowded market | Supply is well ahead of demand. Expect price pressure and a slow start. |
| Scam risk | The work itself is legitimate, but the category attracts fraud. Vet the people you deal with, and never pay to be given work. |
| Income varies | Earnings swing sharply from month to month, or the work is heavily seasonal. |
| Growing demand | Demand is currently outpacing the supply of providers. |
Where this data is weakest
Anything claiming to price a thousand different activities is going to be wrong in places. These are the failure modes we know about:
- It is US-centric. Dollar figures, licensing notes and platform availability assume the United States. Rates in the UK, Canada, Australia and the EU differ, sometimes by a lot, and licensing rules differ enormously.
- Local markets vary more than any national figure can capture. A lawn care route in a dense suburb and one in a rural county are not the same business. Treat local service figures as a starting point for your own research.
- Ranges hide the distribution. Most of these are not evenly spread. Earnings typically cluster near the bottom with a long tail at the top. A range of $0–$5,000 does not mean $2,500 is typical. It usually means a few hundred is typical.
- Fast-moving categories drift. Anything touching AI, platform algorithms or search traffic can change materially within months. The stamped review date tells you how stale a figure might be.
- Survivorship bias is everywhere in the source material. People who quit rarely publish what they earned. We correct for this by weighting the bottom of ranges conservatively, but we cannot eliminate it.
- Difficulty is a judgment, not a measurement. We apply it as consistently as we can, but it is not derived from anything you could independently verify.
What commercial relationships do and do not affect
One entry in this index, KashKick, is a paid affiliate placement. It appears in a labeled featured slot on the homepage and carries a "Featured" tag in the results.
It also happens to rank first on the ease-of-entry score, and it's fair to be suspicious of that. So here's the check. The score comes from the formula above, using data that ships in a file anyone can read. KashKick scores 96 because it costs nothing, needs no skills, and pays out within days. The build process contains an assertion that fails if the featured entry does not win that sort outright on its own numbers. If we ever wanted to prop it up, we would have to change the data itself, and you can read the data. Its earnings figures ($30–$150/month) are also among the lowest in the entire index. That is why it appears near the bottom of every earnings sort.
Commercial relationships do not affect any entry's cost, earnings, difficulty, timing, flags or ordering under any sort. Full disclosure here.
Found something wrong?
If a figure looks wrong to you, we want to hear it, especially if you do that work. Practitioner corrections are the most valuable input this project gets. Send the entry name, what you think the number should be, and roughly what you are basing it on. See the corrections policy for how these are handled and what gets published.