thinverAI · Methodology
How thinverAI evaluates ideas
thinverAI checks business and technology ideas against 11 fixed analysis dimensions: Executive summary, Target customers, Differentiation, Market potential, Competition, SWOT, Feasibility, ROI scenarios, Scaling, Risks, Go-to-market. These dimensions produce five individual scores, an overall score and a Go, No-Go or Pivot recommendation. The result is a structured basis for decisions — not a forecast and no guarantee of success.
Which inputs are processed
Every analysis starts with a free-text description of the project. Optionally, you can add documents whose text content is evaluated as well:
- ›Free-text description of the idea, technology or product
- ›PDF documents, Word, PowerPoint and Excel files
- ›Images such as sketches, diagrams or screenshots
- ›Maximum 25 MB per file
The more precisely you describe the target group, unique selling point, technology readiness and planned business model, the more robust the assessment. If this information is missing, the analysis makes visible assumptions and flags them in the report.
The 11 analysis dimensions
Every report covers the same 11 dimensions:
- ›Executive summary — a short classification of the project
- ›Target customers — primary and secondary target groups and personas
- ›Differentiation — unique selling point, key features, competitive advantage
- ›Market potential — TAM, SAM, SOM, growth rate and possible business models
- ›Competition — main competitors, competitive landscape, market saturation
- ›SWOT — strengths, weaknesses, opportunities, threats
- ›Feasibility — technical complexity, time to MVP, key hurdles
- ›ROI scenarios — conservative, realistic and optimistic, each with margin and break-even
- ›Scaling — scaling factors and expansion paths
- ›Risks — critical and moderate risks plus mitigation approaches
- ›Go-to-market — strategy, channels, initial focus, milestones
The recommendation builds on them as a separate point:
- ›Recommendation — Go, No-Go or Pivot with reasoning and concrete next steps
Market potential: TAM, SAM and SOM
The three market sizes are deliberately shown separately because they answer different questions:
- ›TAM (Total Addressable Market) — the entire theoretically addressable market, regardless of region and ability to deliver.
- ›SAM (Serviceable Available Market) — the part of it that could actually be served with the specific offer in the stated region.
- ›SOM (Serviceable Obtainable Market) — the share that seems reachable within a realistic period.
These figures are estimates based on the description and the model’s knowledge. They don’t replace primary market research and should be checked against official market reports or industry associations before larger investment decisions.
Competition and feasibility
The competitive analysis names relevant providers, assesses market saturation and works out how the project could stand out. Feasibility is assessed via technical complexity, an estimated time frame to the MVP and the key implementation hurdles. The competitors named are a selection, not a complete market overview; market movements after the model’s knowledge cut-off may be missing.
ROI scenarios
Instead of a single revenue forecast, the report shows three scenarios — conservative, realistic and optimistic — each with expected revenue, margin and break-even period. These values are model calculations based on the assumptions described. They are explicitly not a promise of revenue, profit or return on investment.
How the Go / No-Go / Pivot recommendation is made
Five axes are each rated on a scale from 1 to 10: market potential, degree of innovation, scalability, investment appeal and feasibility. Together they form an overall score. The final recommendation sums up the overall picture and always comes with reasons:
- ›GO — the project looks viable; the report names sensible next steps.
- ›PIVOT — the core holds up, but the target group, market or business model should be adjusted.
- ›NO-GO — key hurdles outweigh the rest; the report explains which ones.
The recommendation is an assessment, not a decision. It replaces neither a due diligence review nor your own business judgement.
How assumptions and sources are handled
Market sizes, growth rates and ROI values in the reports are estimates and model calculations, not verified market data. Where information is missing, the analysis makes assumptions and labels them. The example analyses published on this website are also illustrative model calculations for fictional projects, not customer cases.
The AI analysis takes 2–5 minutes. The report costs €49 incl. 19% VAT, one-time.
Limits of the analysis. thinverAI provides automatically generated decision support. The results contain assumptions and uncertainties and are no guarantee of success, revenue or returns. The analysis does not constitute legal, tax or investment advice. An automated patent landscape analysis serves strategic orientation and does not replace a legal freedom-to-operate review or advice from a patent attorney.
Frequently asked questions
- How reliable is an AI idea check?
- The analysis is a structured first assessment, not a forecast. It classifies a project along fixed dimensions and makes assumptions explicit. How robust the result is depends largely on the quality of the input: the more concretely the target group, technology and business model are described, the more precise the assessment. Market figures are estimates and should be checked against primary sources before investment decisions.
- How are TAM, SAM and SOM calculated?
- TAM describes the entire theoretically addressable market, SAM the part of it that can be served with the specific offer in the stated region, and SOM the share that can be reached within a realistic period. thinverAI derives these figures from the project description and the model’s knowledge and labels them as estimates. They are approximations, not the results of primary market research.
- What is the difference between thinverAI and a general AI chat?
- A general chat gives a free-text answer whose scope and structure vary with each request. thinverAI uses a fixed scoring logic: the same dimensions, the same five score axes, the same report structure and an explicit Go, No-Go or Pivot recommendation. This makes results comparable across different ideas, and they are delivered as a full PDF report.
- Does an AI patent analysis replace an FTO review?
- No. An automated patent landscape analysis serves strategic orientation: it shows topic areas, possible applicants and potential gaps. A freedom-to-operate review is a legal assessment of specific property rights in the respective target market and must be carried out by a patent attorney.
- How are submitted ideas and documents protected?
- Inputs and uploaded documents are processed only to create the analysis and are not shared with third parties. Data is stored in the EU (Supabase, Frankfurt). The AI analysis is performed by Anthropic (USA) on the basis of the EU Standard Contractual Clauses. Details on retention periods, legal bases and your rights are in the privacy policy.
Check your idea now
Describe your project in free text — thinverAI turns it into a full PDF report with a Go, No-Go or Pivot recommendation.