A scorer that says no to everything is as useless as one that says yes to everything — both are horoscopes. So we took five ideas we'd already sized up by hand, from obvious duds to genuinely strong, stripped them to a one-line pitch, and fed each to our own free validator cold. We were watching for one failure mode: does the score have real range, or does it hover in a safe, noncommittal 3-to-6 band for everything? Here's exactly what came back.
The spread
1/10 – 7/10
“A social network for everyone to connect with friends.”
Named it instantly — this is Facebook: three billion users, network effects, nothing a new entrant can pry loose. A flat no, and the right one.
Going in, we’d called it a weak idea.
“Dropshipped phone cases from a generic overseas supplier.”
One of the most saturated dropshipping niches there is — no supply edge, no audience, no wedge. Scored it a near-floor 2.
Going in, we’d called it a weak idea.
“A healthy meal-prep delivery subscription.”
A real market, but brutally competitive — HelloFresh, Factor and dozens more already own it. Not dismissed, but you'd need a sharp edge. A cautious middle.
Going in, we’d called it a mediocre idea.
“A tool that pre-checks UK dental practices against CQC inspection criteria.”
Real pain, a real buyer, real regulatory pressure. The top score of the five — and still 'test it first', because the open question is whether practices pay before an inspection or only after.
Going in, we’d called it a strong idea.
“A tool that pre-checks a game against ESRB / PEGI age-rating criteria.”
Didn't reflex-stamp a 7 onto a plausible B2B idea. It found the real bottleneck — the ratings boards control the timing, not the pre-check — and marked it down to a 6 for it.
Going in, we’d called it a strong idea.
Five ideas, five scores from 1 to 7. The thing we were checking for answered itself: the score isn't stuck in a safe middle band. It goes to the floor for ideas that deserve it and climbs for ones that earn it. That range is the whole point — a number with no range is decoration.
Two ideas got a 1 and a 2. This matters more than the high scores. A validator that will never say a flat 'no' is worthless, because the founders who most need to hear it are exactly the ones building a clone or a race to the bottom. This one says no, with a reason, and means it.
The strongest idea — the dental-compliance tool — topped out at 7, not 9. That isn't the model playing it safe. A 9 means validated demand, a defensible moat, and a clear path to scale, and a cold idea on day one structurally has none of that yet. 7 is 'strong foundation' — the honest ceiling an unproven idea can reach. You climb to 8 and 9 by bringing evidence: paying customers, a real waitlist, a moat, then re-running it.
The most telling result was the one that didn't get a 7. The game-rating tool reads like a tidy B2B idea, and a flattering scorer would have rubber-stamped it. Instead the model found the actual flaw — the rating boards, not the founder, control when a game ships — and marked it to a 6. It's reasoning about the specific idea, not pattern-matching 'B2B equals good'.
So the score isn't a pessimism machine and it isn't a cheerleader. It earns its lows and it earns its sevens. The only verdict a cold idea won't get is 'build it' — and that's the honest answer, because an idea no one has paid for yet should hear 'test it first', not 'go all in'.
How we ran this: each idea was written as a single-line pitch and put through the free idea validator — the same public tool anyone can use, no special access, no thumb on the scale. One run each. This is the quick scorer, not the full grounded report, which searches the live web and tends to be even more discriminating. Run the same five pitches yourself and you'll land within a point or so.
Is the Kasspian score biased toward saying no?
No. In this test the scores ranged from 1 to 7 across five ideas — weak ideas scored 1 and 2, a mediocre one scored 4, and strong ones scored 6 and 7. It says no when an idea is a clone or a race to the bottom, but it gives genuine credit to ideas with a real buyer and real pain. A blanket-negative scorer would have clustered everything in a low band; this one didn't.
Why would a strong idea only score 7 out of 10?
Because 7 is the honest ceiling for an idea no one has paid for yet. A score of 9 or 10 means validated demand, a defensible moat, and a clear path to scale — things a cold, day-one idea structurally can't have. 7 means 'strong foundation'. You earn an 8 or 9 by bringing evidence (paying customers, a real waitlist, a moat) and re-running the analysis.
Does the validator ever give high scores?
Yes — the strongest idea here scored 7, and ideas backed by evidence of real demand can score higher. What it won't do is hand out a high score to an unproven idea just because the pitch sounds plausible. The score reflects how strong the idea actually is, not how confident it's described.
Now run yours through the same scorer — same honest read, no thumb on the scale, in about 90 seconds.