Methodology
How we find opportunities, and what the Gold Score means.
How opportunities are found
We look for places where AI tools have recently made something faster, cheaper or newly possible, and where someone is willing to pay for the result. Sources include product announcements, pricing and feature changes from AI tool providers, public case studies, and patterns we see repeated across freelance and small business communities. Where we make a specific factual claim, we link to a source on the opportunity page.
Gold Score
The displayed Gold Scores are version 1 editorial assessments from the original library. Their individual numerical components were not retained, so the totals cannot be independently recalculated. They are not measurements of demand, a verified growth trend or predictions of income. The original rationale considered:
- Demand: how many people or businesses plausibly want this solved
- Momentum: whether interest and enabling tools are growing, steady or fading
- Commercial value: how much a customer could reasonably be charged
- Ease: how quickly someone with modest skills could get a first result
- Competition: how crowded the space already is, which reduces the score
Each opportunity page retains its original rationale, clearly labelled as a legacy assessment. We have removed the unsupported “Rising” trend signal from the public badges. Use the explicit planning thresholds and source checks below to assess a pilot.
Starter plans and the opportunity finder
The finder compares all 15 starter routes. Your chosen problem first narrows the routes. A gap is counted for each threshold you do not meet: implementation experience, starting tool budget and weekly hours. Fewer gaps rank first; an exact experience match breaks ties, followed by library order. When no route fits every limit, the gaps remain visible. Your income, agency or business goal changes the framing, not the ranking. A monthly gross target shows the number of sales or active agreements needed at the guide’s hypothetical fee: target divided by fee, rounded up. It does not prove an achievable price, sales volume or profit.
These thresholds are estimates, not measured success rates, vendor quotes or professional service prices. Recommendations do not depend on sponsorship or affiliate revenue. The comparison page presents legacy Gold Score and practical trade-offs separately. We do not infer numerical sub-scores from the original total. Budget allowances and calculator defaults are hypothetical planning inputs, not researched market fees.
“Content updated” is the stored editorial modification date. It does not certify that every source or vendor price was checked on that date. Where no date is recorded, we say so.
Review history and source types
30 September 2026: added Digital Products and a free BuildKit specification, resource controls on the homepage, suggested buyer and offer-model filters, and gross-target sales scenarios. All routes remain experimental until independently supported sales evidence is recorded. First revenue is not verified. “Low starting allowance” means £100 or less in the guide; “No-code friendly” means implementation level 1. A review date covers the pilot plan, not every linked source.
Version 1 is the original scored opportunity library. Version 2, reviewed on 29 September 2026, adds 14 bounded starter plans, reproducible finder thresholds, downloadable checklists, 19 dated tool pricing records and three attributed external cases. Inspect the research record.
The Ultimate Vision Ltd editorial desk is responsible for the published guidance. AI assisted with drafts; the desk reviewed scope and source descriptions. Provider pages support provider capabilities and prices. Company and vendor case studies report their own results. Neither source type independently proves that your buyer will pay or that your pilot will succeed. Simulations use fictional data and fixed responses.
Interactive samples and cost scenarios
Service demos use fictional documents and fixed local responses. They demonstrate workflow decisions, not a deployed AI system or a customer case study. The cost calculator uses editable hypothetical inputs. It calculates contribution after platform costs and the stated value of support time, before taxes, acquisition costs, initial build labour and other overheads. Its cash-cost recovery estimate assumes a continuing positive contribution; no income is guaranteed.
Sponsorship and affiliate links
Sponsored content is always labelled "Sponsored" and never affects an opportunity's Gold Score. Affiliate links are always labelled "Affiliate link" next to the link itself. See our affiliate disclosure and advertising disclosure for full detail.
What we don't do
We do not publish invented statistics, subscriber counts, testimonials, client logos or income figures. Any number you see on the site is either linked to its source or clearly labelled as an estimate.
Not financial advice. No income is guaranteed. See our full disclaimer.