Glossary

A verdict issued when the Opportunity Score exceeds the BUILD threshold (typically 75/100 or above), the adversary review has been satisfied, and the evidence supports forward investment. A BUILD verdict means the evaluation supports committing product capacity to this opportunity. It is a thesis rating, not a financial guarantee or revenue forecast.

A verdict issued when the Opportunity Score falls in the conditional range (typically 45–74/100) and adversary risks are present but not decisive. A WATCH verdict means the opportunity is real but not yet clear enough to justify a build commitment. The memo will include specific monitoring triggers that, if observed, would support re-evaluating for a BUILD verdict.

A verdict issued when the Opportunity Score falls below the kill threshold (typically below 45/100) and/or adversary evidence is decisive: the structural case against the opportunity outweighs the case for it. A KILL verdict is not a permanent ban on a category. It means the opportunity does not justify commitment at the current time, based on current evidence.

A composite score from 0 to 100, produced by weighting and aggregating the six decision criteria: Search Accessibility, Monetization Signal, Incumbent Defensibility, Category Saturation, Trend Trajectory, and Execution Fit. The score is not a simple average. Criteria are weighted based on their importance to the specific opportunity and market context. A score alone does not determine the verdict; adversary review findings can override a high score.

The classification of an evidence finding based on how it was obtained. AppSniper uses four tiers: T1 Raw (directly observed App Store data — rankings, reviews, keyword results), T2 Derived Fact (deterministically computed from Tier 1 inputs — rank velocity, modernization decay, neighbor role), T3 Inferred Signal (pattern-matched from indirect signals — organic confidence, complaint severity, willingness to pay), and T4 Interpretation (the synthesized read — market structure, primary blocker, primary supporting claim). Every finding in a Decision Memo is tagged with its tier.

A mandatory step in the AppSniper evaluation process where the strongest possible case against the opportunity is constructed. The adversary review covers platform policy risk, incumbent counter-move probability, market timing risk, and data quality caveats. A BUILD verdict cannot be issued without a satisfactory adversary review. The adversary case must be assessed and found insufficient to block the opportunity before a BUILD can be issued.

A decision checkpoint at which the verdict determines whether a project advances to the next stage. In a studio or portfolio context, an AppSniper audit functions as the pre-build gate: concepts must pass evaluation before engineering capacity is allocated. The gate is designed to prevent the common failure mode of building first and validating second.

The ongoing process of comparing AppSniper verdicts against real-world outcomes to improve evidence weighting and threshold accuracy over time. When a team launches a BUILD, AppSniper tracks the outcome and updates the calibration model. Over time, verdict accuracy becomes measurable. That is what separates a real evaluation from an opinion.

The primary output artifact of an AppSniper Opportunity Audit. A structured document containing: a unique memo ID, the evaluation thesis, full evidence stack (each item tier-tagged and scored), adversary review section, Opportunity Score, verdict (BUILD / WATCH / KILL), confidence level, and Next Action Protocol. Decision Memos are designed to be shared with team members, boards, and investors as a defensible record of pre-build diligence.

The specific claim about an opportunity that the evaluation process is designed to check. The thesis is established before any evidence is collected. A well-formed thesis specifies: the target category, the proposed positioning, your proposed distribution advantage, and the success condition. Evaluating without a defined thesis produces unfocused analysis.

Decision Criterion C1. Measures whether new entrants can acquire users organically through App Store search without extraordinary investment. Scored on keyword difficulty, non-branded term volume, brand lock-in degree, and category CPI proxy. A HIGH rating means organic discovery is structurally viable for a new app.

Decision Criterion C3. Measures how entrenched the leading competitors are in a category. Scored on review moat depth (review count × average rating), brand term keyword ownership, update cadence, and switching cost indicators. A LOW defensibility rating indicates incumbents are vulnerable to displacement by a well-executed new entrant.

Decision Criterion C6. The only team-specific criterion in the six. Scores how well your existing distribution capabilities, team expertise, and portfolio cross-sell potential align with the requirements for success in this specific category. The same category opportunity will produce different Execution Fit scores for different teams.

The competitive advantage an incumbent app holds by virtue of a large, high-quality review base. A review moat creates a trust signal that new entrants cannot replicate organically in the short term. AppSniper quantifies review moat depth as part of Incumbent Defensibility scoring. A deep review moat is one of the most reliable WATCH or KILL signals in the framework.

The closing section of every Decision Memo. Specifies the exact recommended next steps based on the verdict. For BUILD: when to begin and what conditions to monitor. For WATCH: what specific signals to track and what thresholds would trigger re-evaluation. For KILL: what structural conditions would need to change for the category to become viable, if ever.

The full set of apps AppSniper considers when evaluating an opportunity: chart-ranked incumbents in the category, unioned with every app that surfaces in App Store search across the opportunity's top keywords (up to 25 results across up to 5 keywords). Built and scored fresh for every opportunity, not reused across categories. A single obvious competitor is not a market neighborhood; AppSniper's is deliberately wider than that.

The classification assigned to every app in a Market Neighborhood: Trigger App, Direct Incumbent, Copycat, Adjacent Substitute, Behemoth, Small Entrant, Stale Incumbent, or Unrelated (Excluded). Roles are assigned deterministically from developer concentration, rating count, modernization decay, update recency, release date, and rank source — not by manual tagging.

The roll-up of a Market Neighborhood's role distribution into one of nine structural reads: Open Niche, Wedgeable but Crowded, Behemoth Fortress, Copycat Swamp, High Demand / Low Supply, Low Supply / Unproven Demand, High Demand / High Supply, Semantic Drift, or Insufficient Evidence. This verdict, alongside the six decision criteria, is what a BUILD, WATCH, or KILL call is actually evaluated against.

A status carried alongside every verdict that answers a different question than BUILD/WATCH/KILL: is the underlying data fresh and complete enough to act on right now? Five states: Build Lockable, Review Required (a spot check is needed first), Watch Only, Kill Lockable, or Not Lockable (data is too stale to trust any verdict, regardless of what the verdict says). A WATCH verdict is always Watch Only; a KILL is Kill Lockable unless the data behind it is itself stale.

Two fields attached to every verdict. The Primary Blocker names the single highest-priority failing gate standing between the current verdict and a better one. The Primary Supporting Claim names the single highest-priority passing gate the current verdict actually rests on. Both are derived from the same gate-priority list, so you can see in one line what's carrying the thesis and what's holding it back.

The confidence status of an individual claim inside the evidence stack, independent of its evidence tier: Confirmed, Partial, Inferred, Proxy Only, Missing, Stale, Contradicted, or Unsupported. A claim's state is resolved against the trust tier of the evidence section it came from, so a Confirmed claim built on Tier 1 raw data reads differently from a Confirmed claim resting on Tier 3 inference.

A hard-fail gate that checks whether the keyword actually driving an opportunity's score still maps to the category being scored. If the top-scoring keyword's vocabulary has drifted away from the category name — a common failure mode in keyword-driven discovery — the verdict is blocked from reaching BUILD until that's resolved, rather than silently scoring a mismatched category as a strong opportunity.

A group of semantically similar user complaints found across one or more apps in a category, with an affected-app count and representative review snippets attached. Complaint clusters that recur across multiple apps in the same Market Neighborhood — not just one app's reviews — are the strongest evidence that a category-wide pain point exists rather than one app's execution problem.

The first hard checkpoint an opportunity must clear before it is eligible for a BUILD verdict at all: does real, accessible search demand exist for the category's keywords, not just chart presence. An opportunity that fails Gate 1 is routed straight to WATCH or KILL — no amount of strong monetization or weak incumbents can override a failed Gate 1.

A check on whether a scoped v1 is actually shippable by a small team in a bounded window, independent of whether the market opportunity itself scores well. An opportunity can pass every market-side criterion and still fail the buildability filter if the thesis requires licensed data, regulatory approval, or platform capabilities outside a lean team's reach.

A named, specific condition attached to a WATCH verdict that would justify re-evaluating the opportunity for BUILD — for example, an incumbent's review velocity dropping below a threshold, or a keyword's difficulty score falling below a target. WATCH is never open-ended; every WATCH memo names the trigger conditions that would move it forward.

A record comparing an original verdict against its real-world outcome once one is known: did a BUILD launch and succeed, did a KILL turn out to be the right call, was a WATCH trigger later confirmed. Calibration entries are how AppSniper's evidence-tier weighting and thresholds are checked against reality instead of staying fixed opinions.

A scheduled follow-up point after a verdict is issued — for a BUILD, typically 60 to 90 days post-launch — where the original thesis is checked against what actually happened in the category. Outcome checkpoints feed the calibration loop and are what separates a one-time verdict from a tracked decision.

The documented record behind every KILL verdict: which evidence was decisive, what the adversary review found, and what structural conditions would need to change for the category to become viable again, if ever. A KILL is not a shrug — it is a specific, revisitable "no" with its reasoning on record.