WantSignal Frequently Asked Questions

A signal is an observation from a source such as search, video or marketplace data tied to an entity and geography. It is evidence of activity, not proof of a sale or future purchase.

Attention is valuable, but it can overwhelm direct-demand charts because most trend observations first tell us what people are noticing. WantSignal therefore gives Awareness its own explorer and map. The default Signals explorer, city/country demand dashboards and primary demand map exclude AWARENESS so attention volume cannot masquerade as commercial demand.

Awareness means people are paying attention. Commercial intent means the term itself has evidence of a product, service, comparison, deal, ticket, merchandise or purchase-oriented need. A sports team, athlete, actor, musician, public figure or cultural event can be strong Awareness without being a shopping signal.

WantSignal resolves known sports teams/leagues and named people through entity registries and aliases. Bare entity names such as a team, athlete or celebrity are treated as informational/awareness evidence by default. Explicit commerce context can change the interpretation: for example, a team name alone remains Awareness, while that team plus “jersey,” “tickets” or “merchandise” can qualify as commercial intent.

WantSignal uses a multidimensional taxonomy rather than a single category label. A normalized observation can carry a category and subcategory, taxonomy path, intent type, commercial-intent level, specificity, classification confidence and taxonomy version while preserving the original source signal class. Ambiguous terms can remain unclassified instead of being forced into a misleading bucket.

A bare platform destination such as “Google” is treated as navigation/awareness with low commercial intent, not as proof that someone wants to buy a Google product. A qualified term such as “Google Pixel,” “Google Cloud,” a product comparison or a purchase-oriented query is evaluated separately.

WantSignal uses curated brand names plus a registry that can discover additional candidates. External category membership can make a name a candidate, but it does not automatically make that brand trusted. Brands gain stronger standing when they are curated or independently observed by WantSignal sources.

The taxonomy covers consumer electronics, home, fashion, beauty/personal care, sports/outdoors, automotive, pets, grocery, books/publishing, music, video games, software/SaaS, artificial intelligence, creator tools, telecommunications, financial services, employment/careers, education, travel, restaurants/foodservice, subscriptions, gifts/deals, seasonal demand, real estate, business/industrial demand, professional/local services, events/experiences, weddings/life events, maker/craft activity, collectibles, entertainment/public figures and other major domains. Recurring unresolved terms are candidates for taxonomy expansion rather than permanent use of a generic “Other” bucket.

No. A source observation can remain AWARENESS or SEARCH even after WantSignal recognizes a team, athlete, celebrity, brand, product family, category or intent pattern. Enrichment adds context around the observation; it does not rewrite the source evidence to make the signal look more commercial than it was.

Raw source-native captures can be stored before classification, but promotion into normalized WantSignal observations is expected to use the shared taxonomy gate so new collectors do not invent incompatible category systems.

For US Designated Market Area (DMA) data, WantSignal requires provider geography provenance tying the stored market to Google's DMA identifier. Repeated normalized values are acceptable when each observation has a validated provider geography contract.

International ADMIN1 observations must carry a country code, provider region code, provider region name and ADMIN1 localization marker consistent with the stored WantSignal geography. This lets WantSignal distinguish legitimate repeated regional evidence from accidental fan-out.

Regional interest is not binary. A term can legitimately appear in many provider-localized regions. WantSignal validates provenance instead of treating repetition alone as corruption.

Geographic strength is the source-specific magnitude of a signal in a market. It is separate from commercial relevance, entity confidence and category context.

It is WantSignal's commercial relevance for the term or commerce relationship, usually expressed as a percentage. It is not the same as geographic signal strength and should not be read as a probability of purchase.

Freshness depends on the upstream source and collector schedule. Each published record includes its observation time, and the UI formats that timestamp in the viewer's locale.

Yes at the data and transport layers. The public API uses UTF-8/utf8mb4, JSON preserves Unicode, and term elements allow automatic text direction. Entity aliases and taxonomy coverage are expanded incrementally across languages and markets.

No. Raw observations, debugging information and research history stay in the research system. The public site receives a curated publication layer.

Yes, when a sufficiently specific and relevant commerce interpretation is available. Marketplace offers are treated as a separate monetization and product-discovery layer. A shopping link or product card does not change the underlying signal into proof of purchase intent.

Coverage remains uneven across sources and countries; taxonomy and commercial classification still include rules-based logic; named-entity registries are not exhaustive; historical depth is still accumulating; and attention does not always imply demand. Those limitations are reasons to inspect multiple signals, sources and geographies rather than rely on a single metric.