Analyzers
Run gscdump analyze <tool> --help for each Analyzer's options.
Thirteen Analyzers support live rows; all 30 support SQL over stored data.
Thirteen Analyzers accept live rows: brand, cannibalization, clustering, concentration, data-detail, data-query, decay, movers, opportunity, seasonality, striking-distance, trajectory, and zero-click. The other Analyzers need a Store.
| Analyzer | What it finds |
|---|---|
striking-distance | Queries in positions 4 to 20 with low CTR |
opportunity | High-impression, low-CTR pages worth optimizing |
movers | Largest changes in clicks and impressions since the comparison period |
decay | Pages losing traffic over time |
survival | How long pages keep receiving traffic |
change-point | Statistical breakpoints in traffic |
brand | Share of clicks from brand and non-brand queries |
cannibalization | Multiple pages competing for the same query |
clustering | Groups of related queries by prefix or intent |
concentration | How traffic concentrates across pages/queries |
seasonality | Weekly and monthly traffic patterns |
stl-decompose | Traffic split into trend, seasonal changes, and remaining variation |
trajectory | The shape of the whole record: peak week, latest week, and a tagged classification |
zero-click | Low-CTR search queries with high impressions and strong positions; clicks can be nonzero |
trends | Clicks and impressions across rolling date windows |
ctr-curve | Observed CTR at each position for the Site |
ctr-anomaly | Pages with CTR above or below the expected rate |
bayesian-ctr | Shrinkage-adjusted CTR estimates |
position-distribution | Ranking positions for each page or query |
position-volatility | How much rankings change over time |
intent-atlas | Search queries grouped by shared words; this does not classify Search Intent |
query-migration | Queries shifting between pages over time |
keyword-breadth | Distinct reported search queries per page with impressions |
long-tail | Traffic split between common and less frequent queries |
dark-traffic | Page clicks minus clicks attributed to reported search queries; an aggregation gap, not a traffic source |
device-gap | Desktop vs. mobile performance gaps |
bipartite-pagerank | Pages and queries with the most influence in their shared graph |
content-velocity | Search queries first observed within the selected window, grouped by week |
data-detail / data-query | Rows for closer inspection |
Run an Analyzer
gscdump analyze <tool> runs one of 30 Analyzers from @gscdump/analysis.
Use gscdump analyze <tool> --help for its flags. The live-capable list above determines whether --live can run it. A stored-only Analyzer needs a complete Store window.
Each Analyzer accepts --site, --period, --start, --end, --limit, --fetch-budget, and output flags.
Windows default to the last 28 days that end on the newest synced day.
trajectory defaults to the last 486 days, the span Google keeps. Pass --start to read a longer preserved record.
trajectory reads impressions by default, and reads clicks when impressions and clicks disagree or the impressions peak falls in the period Search Console over-counted impressions (2025-05-13 to 2026-04-27).
trajectory reads impressions when the peak week has under 20 clicks, and clicks when the peak week has under 200 impressions.
trajectory counts a day without a row as zero traffic up to the end date, and never reads days after it. One outlier day is capped at 3 times the median of the days around it.
movers and decay compare with the previous period by default. Override it with --prev-start and --prev-end together.
movers also accepts --sort-by: clicksDelta (default) sorts by absolute click change, clicksDeltaPercent by percent change.
--limit caps the rows shown. --fetch-budget caps the rows each live fetch reads (default 25000, max 100000).
If a fetch reaches the budget, the output shows a partial-data warning.
Use gscdump analyze <tool> --help for additional options.
analyze and report follow the routing rules.
Pass --live to use Google explicitly.
SQL-only Analyzers require local data.