Where The Data Comes From
Our research draws on publicly accessible social-network content, account relationships, market discussion, and verifiable market data.
- We study public account relationships, historical discussion, and $TICKER mentions across social networks.
- Discussion intensity, engagement, context, and public market information form traceable research leads.
- Every summary should lead back to source material; AI-generated text is never treated as the primary fact.
How We Study Social Networks
Our proprietary community-discovery and propagation research systems identify durable groups, information flow, and meaningful changes in the network.
- Discover related research communities from relationships and public discussion.
- Study how topics emerge, travel, or diverge across communities.
- Use multiple research signals to reduce casual, promotional, and noisy content. Specific models, features, and weights are not disclosed.
- "Trusted author" means one thing everywhere on the site: the account passed an evidence-backed, per-account review as a reliable financial source. Only those accounts count toward author tallies, heat, and sourcing. Reviews are versioned, expire, and defer to human judgment; ranking scores affect order, never eligibility.
How Track Records Are Calculated
Track records only use verifiable $TICKER views, measured by excess return versus the broad market.
- Sample: public $TICKER views from the author. Larger samples are more informative.
- Benchmark: excess return versus the broad market; beating the market means beating that benchmark.
- Holding period: measured after a fixed number of trading days, shown on each author card.
- Small samples are unstable; pages show sample size and distribution context.
Limits To Know
This is a research-sorting tool, not investment advice. AI summaries should be checked against the source posts.
- Follow-graph updates can lag by about a week.
- Track-record samples can be thin; past performance does not predict future results.
- Account types are AI-classified and may be wrong.
- Community boundaries are algorithmic; edge accounts may move between groups.
- Digests and search answers are AI-generated; source posts are the ground truth.
Research First, Kept Accessible
Roughly 80% of SpikePanel is devoted to social-network and information-flow research, with the other 20% focused on turning that work into a useful product. We are not building a high-priced information gate. Our aim is to make social research leads worth checking accessible at a low cost. Paid access primarily supports AI inference and the ongoing work behind the research, while we keep pricing aligned with the value delivered and as accessible as possible.