Trend Discovery Configuration Guide

What Discovery Does

The Discovery Engine reads trend signals for each client account's topics and feeds every signal into a scoring pipeline tuned to that account. Each raw signal (a trending search, a new story about a topic) is evaluated against four weighted criteria: trend velocity, brand relevance, competition gap, and historical performance. The highest-scoring signals become content opportunities you can send to production.

Trend Sources

Discovery reads these sources on every scan. There is nothing to add or configure per source:

• Google Trends, the day's trending searches for the configured region.
• Google News, recent news stories for each of the account's topics.
• DataForSEO, Google Trends interest over time and rising related queries for the account's topics, where DataForSEO is enabled for your workspace.

Social signals, competitor monitoring and seasonal data are planned and not live yet.

Choosing Topics

Discovery searches for the account's topics. It builds them from two places in the client account, so set them up in Buckets and Personas before the first scan:

• Content buckets: each bucket name is a topic (e.g. remote work culture).
• Persona interests: the interests listed on each persona are added as topics.

A scan uses a limited number of topics per account, bucket names first, so keep bucket names specific.

Scoring

Every signal is scored against four criteria. The weights are fixed and the same for every account:

• Trend velocity, 30%. How fast the topic is growing in the source data.
• Brand relevance, 30%. How closely the topic matches the account's persona interests.
• Competition gap, 20%. Neutral today: competitor monitoring is planned, so every topic gets the same middle score.
• Historical performance, 20%. How similar topics and sources performed for this account in past batches.

Opportunities are listed by score on the Discovery page.

Prohibited Topics

Topics listed as prohibited on the client's brand are dropped before an opportunity is created. Set them in the brand library; there are no separate Discovery filters.

When Scans Run

A scan runs when someone on the team starts it:

1. Open Discovery in the client account.
2. Click Scan for opportunities. The page shows the scan as queued, then running.
3. When the scan completes, new opportunities appear in the list with their scores.

Scheduled scans are planned. Until then, run a scan whenever you plan a new batch.

Trend Scout Agent

The Trend Scout is the AI agent that runs on top of the Discovery Engine and synthesizes cross-source signals into a ranked opportunity brief. It runs after every aggregation cycle, groups related signals by theme, and writes a short rationale for each top-ranked topic. See Platform Overview for how Trend Scout fits into the full content pipeline.

Common Gotchas

• Empty discovery results. A scan with no topics has nothing to search for. Add content buckets or persona interests to the account, then scan again.
• Trending vs. evergreen tradeoff. Velocity carries the most weight with relevance, so fast-moving topics rank first. If a client needs a steady baseline of evergreen content, keep those themes as content buckets and pick them from the list alongside the trending ones.
• Region is workspace-wide. Google Trends and, where enabled, DataForSEO read the region configured for your workspace (United States by default), not a per-account setting.