Skip to content
Courses / Will AI Eat This Business? / Score your own idea

Score your own idea

5 min read · Your assessment

Lesson video in production

The full lesson text below is complete — the video version lands with launch.

Score your own idea

The method only counts if it travels outside this catalog. So before you score your own idea, watch it score two it does not contain. I picked one idea the scorecard should reject and one it should bless, and I will run both in the open, arithmetic visible, so you can check my work line by line.

First run: a resume-writing service

Somebody is selling this idea online right now, and it sounds plausible: people hate writing resumes, job changers pay for polish, start-up cost is a laptop. Start at 4.

Deliverable. A resume is bytes. Formatting, bullet points, cover letters, LinkedIn rewrites: all of it is exactly the material chatbots produce acceptably, and job seekers have noticed. Add 2. Running total: 6.

Entry floor. The simple version of this work three years ago was polishing a resume for a career changer, a few hundred dollars a client, and that version now generates for free. The remaining human rung, executive career coaching with placement insight, is a real business, but it is a different business a beginner cannot yet sell. Add 2. Running total: 8.

Demand gate. Job seekers find resume writers through search and gig platforms, mostly unpoliced ones. Add 1 for the platform exposure, being fair, since some demand flows through referrals. Running total: 9.

Blame. A resume writer is rarely answerable for outcomes; no one guarantees the interview. Nothing to subtract. Running total: 9.

Task split. Most of the year's hours sit on writing and rewriting documents, machine-doable. Add 1, then clamp. Final score: 10 is too confident for a service where interview coaching and veteran niches survive, so I read this honestly as a 9, with a residual lane for people already established career coaches. A beginner entering today is entering head-to-head against a free tool that is adequate at the exact thing they would charge for.

The verdict: do not build this. That will sting if you were considering it, and I would rather it sting here than in month four.

Second run: a professional home-organizing service

Garages, closets, pantries, playrooms. Also a real business people run, and mostly absent from the hustle-content internet. Start at 4.

Deliverable. An organized home is a physical outcome, produced in the customer's house over a day or two. Subtract 2. Running total: 2.

Entry floor. The simple version three years ago was helping someone declutter a closet. It still requires a body, a van for donations, and a personality that can ask a stranger to throw away their grandmother's casserole dishes. The floor holds. Nothing to add. Running total: 2.

Demand gate. Organizers are found through word of mouth, neighborhood groups, and local search for a physical service, which behaves like referral demand, not informational search. Subtract 1. Running total: 1.

Blame. The organizer is alone in the home with the customer's possessions, trusted like a cleaner, sometimes bonded. Real accountability. Subtract 1 more, honestly, for the trust requirement. Running total: 0.

Task split. Production is sorting, lifting, binning, and coaching a customer through decisions. Almost nothing a machine can do without you. Hold. Final score: 0 to 1.

The verdict: build it. The risks in that business are the ordinary ones, pricing, scope creep, the emotional labor of decluttering, and none of them are a model release.

Your run

Same discipline, your idea. One hour, one page, no shortcuts.

Your one-hour assessment

Write the ten-task inventory and mark each R or A
Name the deliverable: bytes, outcome, or accountability, and strip the bundle if it is mixed
Find the demand gate and check whether its gatekeeper polices machine supply
Run the entry-floor test on the three-years-ago version of the simplest work
Start at 4, apply the adjustments, clamp, and write the number down
Have one blunt friend run the same rubric on your idea and compare numbers

Then act on the band, not on hope. A 0 to 3 means the catalog almost certainly has a course for it, and the ordinary risks deserve your attention instead. A 4 to 6 means build, with the four adaptation moves from the previous lesson designed in from the start. A 7 or above means pick another idea. It is not weakness to walk away from a closed door; it is the entire point of doing the assessment before you spent the savings.

One more habit worth stealing from how we run the catalog: date your score. Write the number, the date, and the three facts that drove it. When a trigger fires later, you will want to know what changed, them or you.

Last stop: how this catalog was actually made, and where to go next now that you can read its numbers.

Downloads for this lesson

Keep going — you're working through Will AI Eat This Business?.

All courses are free ↗