Begin with the source
Spectre brings market information into a shared research workspace. Data delivered through Spectre can originate from an external provider, an exchange, a published announcement or social posts. Check the source label or linked material available in the view.
Source attribution matters because datasets can measure different things. The public tokenized-assets chart, for example, credits DeFiLlama data via Spectre and describes tracked total value locked in US dollars, excluding stablecoins. That measure is different from a token’s market capitalization or trading volume.
Check the clock and the comparison
The time an event happened, the period covered by a metric and the time you opened a page are different. A recently loaded page may contain an observation reported earlier. Use the displayed date, period and source context; do not assume every panel updates on the same schedule.
The public economic calendar displays times in UTC and uses a dash when a value has not been reported. Treat missing or unavailable data as unknown, rather than zero. A chart gap is not evidence that no market activity occurred.
- Match the asset and, where relevant, its chain or contract.
- Check the unit: dollars, tokens, a percentage, a count or an index.
- Use comparable windows and note whether an equity session was open or closed.
- Check the latest available observation before calling a movement current.
Separate observations from interpretation
A price change, published number or source post is evidence to examine. An AI summary connects observations and suggests an interpretation. Read the supporting material, especially when a conclusion depends on one event or a disputed claim.
A good comparison also considers another explanation. Rising attention and a rising price may occur together because both followed an announcement. Their overlap does not establish which caused the other. A historical pattern is not a promise about the next move.
Understand coverage and visual choices
A selected set of projects or posts is not the whole market. Sources can differ in asset coverage, definitions and reporting intervals. Compare results within the same scope before treating a difference as a meaningful signal.
Visual size and colour need their own context. Spectre’s public market bubbles use a compressed size scale, so a bubble twice as wide does not imply twice the market capitalization. In social research, activity and engagement do not automatically establish positive sentiment or credibility.
Keep a research trail. Report a discrepancy.
When sharing a finding, include the asset, metric, period, source link and the observation you used. Keep the evidence separate from your conclusion so another reader can follow the reasoning.
If a Spectre page appears incorrect, contact spectre@spectreai.io with its URL, the affected metric or passage, the time you observed it and a supporting source if available. Avoid sending passwords, access keys or private account information.
Product screenshots show historical examples. Beta features and available data can vary by access level. This guide explains a research workflow; it does not provide a trade recommendation.