What is the studio actually good at?

Quality Analysis: What the Studio Builds Well and Where It Fails

The honest picture

The mean composite of 68.2 across 822 products sounds passable until you look at the actual distribution. 808 of 822 products (98.3%) are smb-saas, averaging 68.0 with a range of 32 to 98. That 66-point spread is the whole story. The other 14 products (enterprise, side-hustle, lemonade, mid-market) average well above 74, but they are statistical noise at that sample size, not actionable patterns. The real question is what separates the 32s from the 98s inside the smb-saas bucket, because that is where all the volume lives.

What scores high and why

The top-performing verticals are legal (85.5), accounting (85.0), and trucking (81.3). These are not accidents. Legal and accounting are rule-bound, document-heavy industries with a short list of universally understood pain points: contracts, filings, compliance deadlines, billable hour tracking. When a product addresses one of those, the landing page writes itself. The value proposition is concrete, the buyer is trained to pay for tools, and the category framing maps cleanly to established software categories buyers already search for.

Trucking scores high for the same structural reason with different surface area. Dispatch, DOT compliance, load tracking, driver logs: the vocabulary is specific and the pain is operational and daily. The studio articulates operational specificity well.

Top categories by average: intake (80.7), analytics (78.5), finance (78.0), document-automation (76.0). These share one trait: a clear, auditable output. A document was produced. A number was calculated. A form was completed. Buyers in these categories evaluate tools on whether the output exists and is correct. That is a solvable framing problem.

What scores low and why

The composite minimum is 32, entirely within the smb-saas and smb-human pools. Every named category averages above 73, which means the products dragging the mean to 68.2 live in categories the report does not surface, or they represent the generic tail of smb-saas. The content category at 73.0 (lowest named) is the early warning sign: content tools are commoditized, the buyer's expected outcome is vague, and differentiation is nearly impossible to articulate without tight niche context.

Products below 40 are almost certainly built on fuzzy briefs: "AI tool for small businesses," "productivity app for teams," "marketing assistant for entrepreneurs." The studio cannot write a sharp landing page for a vague product. It is producing the output correctly. The input is underspecified.

The underlying pattern

Specificity of buyer pain correlates directly with quality score. Legal and accounting buyers have highly specific, recurring, expensive pain. A product that solves one named problem in one named vertical scores well because every sentence can anchor to something real. Generic SMB tools have no such anchor. The vertical and category signals in this data are really proxies for brief quality, not for inherent product difficulty.

Concrete recommendations

• Bias heavily toward verticals with regulatory overhead: legal, accounting, medical, construction, trucking. Compliance burden creates perpetual, articulable pain.
• Weight toward document-automation and finance categories. Clear inputs, clear outputs, buyers who already have software budgets.
• Within ops-tooling (n=26, avg=73.4, the largest real sample outside smb-saas), prioritize products that name a specific role or workflow rather than "operations generally." The range is probably wide; the high end lives in specificity.
• Dial back on generic productivity tools, collaboration platforms, and any product whose description could apply to 50 existing SaaS products. If the idea would not survive a LinkedIn ad targeting a specific job title in a specific industry, skip it at the mining stage.
• Treat verticals with fewer than 5 products as hypothesis only. Legal at 85.5 across 4 products is a promising signal, not a confirmed pattern.

The controversial take

The enterprise product scoring 96 and the side-hustle tier averaging 79.2 probably score higher not because those products are better designed, but because the brief format for those tiers forces specificity by definition. "Enterprise" requires naming a specific large-company workflow. "Side-hustle" requires naming a specific service someone sells to people they know. The smb-saas tier fails as a category because it is a catch-all for vagueness. If the miner used effort-tier framing as a brief-quality filter instead of a market-size signal, scores in the smb-saas bucket would probably rise without changing any other part of the pipeline.