Methodology

Methodology You Can Understand, Question, and Trust

SharedToday is designed to make complex digital discourse measurable and useful while remaining transparent about what the data means, what it does not mean, and how it should be used responsibly.

What SharedToday Measures

SharedToday is a social and public discourse intelligence system. It is not a polling system, not a census, and not a user-level geolocation system. Its role is to detect patterns in public digital expression across platforms and convert those patterns into structured, decision-ready intelligence.

What SharedToday measures

SharedToday analyzes publicly available digital conversations and related signals to identify issue emergence, topic concentration, sentiment direction, engagement intensity, narrative spread, and comparative discourse patterns.

What SharedToday does not measure

SharedToday does not claim to measure the full population, does not verify user-level geography, and does not treat digital discourse as a direct substitute for polling or demographic research.

Data Sources and Coverage

Depending on project scope, SharedToday may cover social platforms, public pages, public groups, public posts, websites, and selected public profiles where relevant and compliant. Additional internal or partner-provided data sources can be integrated on a project basis to support richer contextual intelligence.

Coverage is configured

SharedToday is designed around analytical need rather than a fixed source template, which helps reduce noise and increase relevance.

Project-specific enrichment

Internal, partner, or structured project datasets can be integrated where needed to support richer contextual interpretation.

How Geography Is Handled

SharedToday does not perform verified individual-level geolocation. Geographic relevance is interpreted through source context, content signals, known mappings, and conversation clustering rather than certainty about the physical location of each user interaction.

Geography is contextual, not certain

Location in SharedToday should be read as an analytical interpretation layer built from context and clustering, not as verified physical user-level geography.

How Topics Are Detected

Topic detection combines NLP-based extraction, phrase clustering, semantic grouping, and frequency or engagement-based ranking. The purpose is to surface recurring and emerging issue clusters in digital discourse.

Topic detection supports issue discovery

The goal is to surface patterns quickly so researchers, analysts, and decision-makers can understand what is clustering and why it may matter.

How Sentiment Is Detected

SharedToday’s sentiment interpretation is text-based. It analyzes posts, comments, captions, and related textual content to identify directional sentiment patterns. It does not assign sentiment from reaction types alone.

Directional, not magical

Sentiment works best as part of a broader analytical picture that includes topic movement, narrative spread, and contextual reading.

Understanding STS and STC

SharedToday uses context-specific scoring approaches to help interpret engagement and comparative discourse behavior. STS is used to understand internal engagement strength or intensity within a defined context. STC is used to compare relative discourse position within a given dataset or competitive environment. These scores should be read as analytical signals, not as direct proxies for population support or final outcomes.

STS

A signal for internal engagement strength or intensity within a defined context.

STC

A comparative signal for relative discourse position inside a given dataset or competitive environment.

How Recommendations Are Generated

Recommendations are derived from observed shifts in topic concentration, sentiment direction, issue momentum, comparative movement, and narrative change. Their purpose is to help users translate signal into action, not to replace human judgment.

Recommendation logic

SharedToday recommendations are there to support decisions, sharpen response planning, and guide analysts toward the next most relevant question.

What to Keep in Mind

Digital discourse is not the whole population. Online activity may overrepresent highly vocal users, reflect platform bias, and include ambiguity in language, sarcasm, and context. SharedToday is strongest when used as a fast, large-scale intelligence layer and, where necessary, complemented by surveys, field research, analyst validation, or domain expertise.

Best used for

issue discovery, narrative monitoring, comparative discourse intelligence, early warning, executive briefing support

Not intended as

a substitute for representative polling, verified demographic attribution, certain user-level geolocation, a stand-alone source for high-stakes conclusions without context

Methodology should strengthen trust, not hide complexity.

If you want to see how the methodology maps onto your own use case, SharedToday can walk you through the framework in a tailored briefing.