Topic Trending

How velocity and volume combine

Volume is the raw count of mentions matching a topic in a given window. Velocity is the rate of change of that count over consecutive windows, often expressed as a multiple of a trailing baseline or as a week-over-week delta. A topic can be high-volume and flat (an established, ongoing conversation), low-volume and rising (an emerging theme), or high-volume and falling (a conversation losing steam).

Trend detection usually ranks topics on both metrics together rather than on either alone. A theme that suddenly doubles from a tiny base is interesting but fragile; a theme that climbs steadily for three weeks from a moderate base is the more reliable emerging-conversation signal. Setting the baseline window matters as much as the math: too short and any off-topic spike looks like a trend, too long and a real shift is reported weeks after it began.

Topic trending vs. conversation clustering and trend analysis

Topic trending overlaps with two adjacent ideas and is often confused with both. Conversation clustering is the upstream step: an algorithm groups unstructured mentions into themes, often without predefining them, which is what produces the set of topics that could then be tracked. Topic trending is the downstream measurement of how those discovered themes, or a manually defined set, move over time.

Trend analysis is the broader, older concept it sits inside. Trend analysis can run on any time series, market or otherwise, and may cover years or decades; topic trending is the application specifically to conversation volume over short windows. Practitioners use clustering to find candidate topics, then trending to track them, and call on the wider trend-analysis toolkit when they need to separate signal from seasonal noise.

How competitive intelligence teams use it

A CI team typically maintains a fixed topic list per competitor and per market category and refreshes velocity weekly. For B2B SaaS, productive topic lists include the names of competitor products, category terms, integration partners, and recurring pain phrases pulled from review sites. A topic that climbs across competitor blogs and release notes in the same window often previews a messaging or roadmap shift before any single announcement makes it obvious.

The same readout also feeds faster decisions. A pricing-related topic rising across competitor pages and review comments in a given week is a cue to check whether a tier change or discount is rolling out. A hiring-topic cluster gaining velocity on a competitor's job board postings can flag an investment area earlier than a press release would. Pairing topic trending with continuous monitoring of competitor websites, pricing pages, and job postings lets the team re-rank the topic list on evidence rather than on what was memorable in the last standup.

Common mistakes and limitations

The most common failure is treating any week-over-week spike as a trend. Single-window jumps are often driven by one viral post, a conference, or a platform algorithm change, and they regress the following week. Requiring velocity to hold across two or more windows, or confirming against a longer baseline, filters most of this noise.

A second failure is topic drift. A keyword-defined topic silently picks up unrelated mentions as language shifts, so a steadily rising curve sometimes reflects a broadening definition rather than a real conversation. Periodically re-clustering the underlying mentions and re-reading the topic definition catches this. Finally, topic trending reports movement, not cause; it tells the team a conversation is accelerating but not why, and the why still requires reading the mentions themselves.

Frequently Asked Questions

Related terms