Every idea in the library has been through the same research template, which is the whole point: they can be compared rather than admired. This page explains what the numbers mean, what things cost, and how to get an idea of your own in.
On the free plan you can see every idea in the library — the titles, the categories, all seven scores — and open one in full each day. It is drawn for you, it is a different one tomorrow, and it is never one you have already kept.
That daily one disappears when tomorrow's is drawn, unless you save it. Saving is free, it is permanent, and it is the single most useful button on the site:
Save keeps an idea for good. The button is on the idea's own page, next to the like and share controls, and on its card in the library. A saved idea stays open to you whatever happens to your plan afterwards — and it is taken out of the daily draw, so saving also stops you being offered the same one twice.
Money is the one thing the free plan sees as ranges rather than exact figures. A market shown as “$1B – $5B” is not the research hedging — there is one exact number behind that band, and Serious prints it. How much to trust the figure is a separate thing entirely, and it has its own label: see confidence, below.
Every idea carries one headline grade, seven scores underneath it, and a confidence label. They answer three different questions and it is worth not confusing them.
A letter from the seven scores below it, weighted. It is a summary, not a verdict — and because the weighting is yours to change, the same idea can be a B+ to someone raising a round and a C+ to someone building alone at weekends. If the grade on a card differs from the one on the idea's page, that is why: the card uses your weighting and the page uses the house one.
Market, feasibility, moat, money, timing, differentiation and speed. Each is out of 100 and each is judged on its own — a high market score says nothing about whether you could actually build it.
This is the honesty signal, and it is not the grade. Low, medium or high says how well-sourced the numbers are, not how good the idea is. A high-confidence C is a weak idea we are fairly sure about; a low-confidence A is a promising one built on thin ground. Read them together or neither means much.
Every market size and unit economic on the site is a modelled estimate from a structured research pass, not audited market data. That line is printed at the foot of every idea, and it is meant literally. The charts and tables under “The figures” show you the same numbers as the tiles above them; where an idea has been through a full research pass, its sources are listed and dated so you can check the ones that matter.
The library has four scopes down the left:
Search reads the title, the problem, the solution and the tags, so you can look for a market rather than a name. The filters below it narrow by category, venture type, business model, how much capital it needs and how confident the research is, and every count updates as you go. The figures are filters too — a minimum score, a minimum market, a ceiling on what it costs to start or how long it takes to build.
Reading as changes the weighting behind the scores, and the library re-ranks live. An investor and someone building solo do not want the same idea at the top. This is kept in your browser, so it does not follow you to another device.
Five views of the same filtered set: the dashboard summarises it, Table puts every figure in one grid, Matrix plots feasibility against market so the top-right is where the easy wins are, and Compare puts a few side by side, field by field.
Credits are what you act with. Reading costs nothing — they are for putting something into the library, or for sending something to another member.
| What | Credits | Who pays whom |
|---|---|---|
| When you sign up | 50 | a one-off welcome grant |
| Every month | 300 | your plan's allowance, topped up |
| Send an idea for review | 30 | spent |
| Keep an idea private | 10 | spent |
| Send a private one for review later | 20 | spent — so 10 + 20 lands in the same place as 30 |
| Revise an idea already published | 10 | spent |
| Have us run the research for you | 500 | spent — writing the prompt yourself is free |
| Like an idea | 1 | from you, to its author |
| Comment on one | 3 | from you, to its author |
| Share one to your profile | 1 | from you, to its author |
Publishing pays no bounty. Getting an idea published earns you nothing at the moment it happens — the credits you spent bought a curator's time. What a published idea earns you comes afterwards, from its readers: every like, comment and share moves credits out of their balance and into yours. An idea that lands keeps paying; one that nobody reads does not.
Saving up 3,000 credits redeems 50% off one month of Serious, applied to your next invoice. It discounts a subscription, so it needs one to apply to.
There are two ways, and the free one is the good one.
The alternative is to have us run the research instead, which costs 500 credits. It is the same prompt. You are paying to skip the copy-and-paste, not to get something better.
Everything you have ever sent for review is listed on your account page under Your contributions, with its reference, what it cost, its state, and the note the curator left. Accepted ones link straight to the published idea.
The published idea carries your byline and appears on your public profile, which counts what you have had published and the likes it has collected. Readers can follow you from it. You can revise a published idea for 10 credits, or close its comments, from the idea's own page.
And it is in the Library now, not in Mine. Mine is drafts.
On the submit page you can keep an idea to yourself instead of sending it for review. It costs 10 credits, it lives in the Mine scope, nobody else can see it, and you can edit it free and as often as you like.
When you are ready, send it for review for 20 more. It keeps the same reference and the same record — the private idea is the thing a curator reads, not a copy of it. If a curator turns it down it goes back to being private, with the reference intact.
One thing to know before you choose: a curator cannot judge a title. If you keep a bare note with no research on it, there is nothing to send for review later — do the research first, then decide where it goes.
The contact form reaches a person who reads every message. If it is about your own data — a copy, a correction, an export, deletion — say so and it is answered within 30 days, as the Privacy Policy commits to. The Terms cover the rest.