Skip to main content
Card class: Sensitivity • Category: Nerve Centre
Catches algo updates and ranking shifts on the queries that drive revenue, before clicks-trend cards do.

​
At a glance

A real-time alert that fires when any query currently ranking in the top 10 falls more than 5 positions against its trailing 7-day average position. A top-10 query slipping to page 2 is the single fastest leading indicator of a Google algorithm update, a content-decay problem, or a competitor overtaking you. Because average position moves before clicks do, this alert leads Clicks Trend by days. The card is a fire-alarm bell for ranking erosion, not a measurement of traffic.
What it tracksPer-query average position for the property’s commercially important queries. For each query that was in the top 10 (average position ≤ 10) over the baseline window, we compare its latest 7-day average position against its prior 7-day average and fire when the position number rises (worsens) by more than 5.
Data sourceGoogle Search Console Search Analytics (Performance report) at the query dimension, read via the Search Analytics API. We pull average position by query and compute the per-query delta.
Why it mattersPositions 1 to 3 capture the bulk of organic clicks; positions 11+ (page 2) earn almost none. A 5-position drop from, say, 4 to 9 roughly halves the clicks a query can earn; a drop from 8 to 13 can wipe them out entirely. Catching the position move early lets you act before the click and revenue loss lands.
Time windowRT (real-time framing). Search Console Performance data is itself delayed (see Reconciling below), so we surface the position drop on the next poll after Google publishes fresh data for the query.
Alert triggerany top-10 query dropping >5 positions vs 7d baseline, sentiment key gsc_ranking_drop. The >5 threshold filters out the normal day-to-day position wobble that affects almost every query.
Rolesowner, marketing

​
Calculation

Calculated automatically from your Google Search Console data. For each query that held an average position of 10 or better over the baseline window, we compare its latest 7-day average position against its prior 7-day average position; the alert fires when the position number increases (worsens) by more than 5. See the At a glance summary above and the worked example below.

​
Worked example

A UK outdoor-gear retailer, Search Console verified on trailgearco.co.uk. The brand watches a basket of high-intent commercial queries. A broad Google core update begins rolling out on 12 May 26.
QueryAvg position (prior 7d)Avg position (latest 7d)ChangeStatus
”waterproof hiking boots”3.24.0+0.8Normal wobble
”lightweight tent 2 person”5.111.4+6.3Alert: dropped 6.3 positions
”merino base layer”4.810.9+6.1Alert: dropped 6.1 positions
”trekking poles”7.08.2+1.2Normal wobble
Three numbered observations:
  1. The pattern points to an algorithm update, not a technical fault. Two queries dropped sharply while the rest held. The affected queries share a content type: long-form buying guides that had thin, templated intros. The core update down-weighted that thin content. Because the URLs are still indexed (no coverage drop on Index-Coverage-Drop Alert), this is a ranking re-evaluation, not a de-indexing.
  2. Triage workflow when this card fires. Order of investigation: (a) check the calendar against Google’s confirmed update history, a cluster of drops dated to a known core or spam update points to algorithmic re-ranking; (b) confirm the affected URLs are still indexed (rule out de-indexing); (c) open the Performance report, filter to each dropped query, and read the “Pages” tab to see which URL ranks now, sometimes Google has swapped which of your pages it ranks, splitting authority; (d) search the query manually (incognito, correct region) and see who overtook you, a new competitor or a freshly updated rival page is a content-quality signal; (e) review the page against the query intent: has the content gone stale, lost depth, or fallen behind competitors?
  3. Why the alert leads the clicks cards. The position drop registered on 14 May; clicks for the two queries did not visibly fall until 17 to 18 May, once the lower positions had been live for several days of impressions. This card flagged the erosion three to four days early, giving the brand time to refresh the two buying guides and request re-indexing before the click loss fully compounded. By 28 May “merino base layer” had recovered to position 6.2 after a content refresh.
Rule of thumb. Several top-10 queries dropping together on the same date = suspect an algorithm update; a single query dropping in isolation = suspect a competitor move or page-specific decay. Either way, the page is still indexed, so the fix is content and relevance, not robots or canonicals.

​
Sibling cards merchants should reference together

CardWhy pair it with Ranking-Drop Alert
Average PositionThe property-wide companion. This alert is per-query; Average Position shows the aggregate so you can tell a localised drop from a site-wide one.
Position TrendThe 30-day position trajectory. Use it to see whether a dropped query is recovering or still sliding.
Queries Dropping From Top 10The list view of exactly which queries have fallen off page 1.
Declining QueriesThe broader set of queries losing ground, not just those that left the top 10.
Ranking VolatilityThe instability gauge. High volatility across many queries on one date is a strong algorithm-update signal.
Index-Coverage-Drop AlertThe de-indexing alarm. Run it first to rule out that the ranking drop is actually pages leaving the index.
Clicks TrendThe lagging-indicator validation. Clicks fall a few days after this card fires.
Position DistributionThe shape of your whole ranking profile, to see how many queries sit near the page-1 cliff.

​
Reconciling against the source

Where to look in Google Search Console:
Performance → Search results. Switch on the “Average position” metric, set the date range to “Compare”, and add the “Query” dimension. Sort by position change to find the queries that slipped. This is the closest native rebuild of what the alert watches. Performance → Pages tab (filtered to a query). Once you have a dropped query, filter to it and open the Pages tab to confirm which URL Google now ranks. A page swap is a common, easily missed cause of a position drop.
Other GSC views that are related but are not this alert:
  • Performance “Position” line on the overview chart: that is the property-wide average across all queries, which can stay flat while individual money queries collapse. This alert is deliberately per-query.
  • Insights: Google’s curated highlights; useful colour, not a precise per-query position diff.
Why our number may legitimately differ from the GSC UI:
ReasonDirection of divergence
Reporting delay. Performance data is typically 2 to 3 days behind. Our latest reading lags Google by the same amount, so a manual check today may not match a freshly published figure.Lag of 2 to 3 days
Average-of-averages. GSC’s “average position” is itself an impression-weighted average across many searches; comparing two 7-day windows can read differently from the UI’s own comparison if the impression mix shifted.Variable
Anonymised queries. Google withholds rare queries to protect user privacy. A query that crosses the anonymisation threshold can appear to “drop in” or “drop out” of the data even though its true ranking did not move.Variable
1,000-row cap. The Performance UI export caps at 1,000 rows. If you reconcile a long-tail query against a capped export, it may simply be missing from the view, not from the alert’s API pull.Query missing from UI export
Position is an average, not a rank check. A query that ranks #2 on desktop and #15 on mobile shows a blended average. A device-mix change can move the average without any single ranking moving.Variable
Cross-connector reconciliation:
CardExpected relationshipWhat divergences mean
google_search_console.clicks-trendClicks for the dropped query fall a few days after the position drop.Clicks falling without a position drop points to a CTR or snippet problem, not ranking.
google_search_console.branded-search-cannibalisation-gsc-vs-google-adsIf a query is also bid on in Google Ads, paid can mask the organic position loss in total traffic.Use it to separate the organic ranking loss from paid coverage.
google_analytics.organic-vs-paid-trafficGA4 organic sessions to the affected landing pages decline after the ranking drop.Steady GA4 organic alongside a position drop suggests the query is low-volume.
This card is not the source of truth for live rankings; Google Search Console is. Average position in GSC is a blended, delayed, sampled figure, not a live SERP check. This card watches that figure for the queries that matter to your revenue and raises the alarm early, so you can investigate inside Google’s reporting window rather than after the clicks have already gone.

​
Known limitations / merchant FAQs

The alert fired but when I search the query I still rank on page 1. Is it wrong? Average position is not a live rank check. It is an impression-weighted average across every search Google logged for that query over the window, including searches from different locations, devices, and personalisation states. Your own incognito search from your office is a single sample. Both can be correct at once. Treat the alert as “this query’s blended position worsened materially” and use the Performance report’s compare view to confirm. Why does it only watch top-10 queries? Because that is where the revenue is and where a drop hurts most. A query falling from position 30 to 40 changes almost nothing; a query falling from 4 to 11 falls off page 1 and loses most of its clicks. Restricting the watch to queries that were in the top 10 keeps the alert focused on losses that matter. Movements deeper in the long tail are tracked by Declining Queries. Several queries dropped on the same day. Is it an algorithm update? A cluster of drops sharing one date is the classic signature of a broad core or spam update. Cross-check the date against Google’s published update history. If the date matches and your indexed-pages count is steady, treat it as algorithmic re-ranking and respond with content quality work, not technical fixes. A query dropped because Google swapped which of my pages ranks. Does the alert understand that? The alert sees the query-level position move; it does not automatically attribute the cause. Page swaps (Google deciding a different URL of yours is more relevant) are a frequent cause and are easy to miss. After the alert, always open the query’s Pages tab in the Performance report to check whether the ranking URL changed. If it did, you may have a keyword-cannibalisation problem to consolidate. How fast does this alert fire? It is bounded by Google’s reporting cadence. Performance data lags 2 to 3 days, so the earliest we can surface a drop is once Google publishes the lower position, typically 2 to 3 days after the SERP actually changed. That still leads the click-loss cards, because clicks take further days to fall after positions slip. Why did a query “drop in” out of nowhere? Google anonymises rare queries to protect privacy. A query hovering near the anonymisation threshold can appear and disappear from the data without its true ranking changing. If an alerting query has very low impressions, treat the position figure with caution; small denominators make average position noisy. Can I tune the 5-position threshold? Yes, sensitivity thresholds are configurable per profile in the Sensitivity tab. Lower it if you want earlier warning on shallow drops; raise it if a volatile niche generates too many low-stakes alerts. The default of >5 positions is chosen to flag a meaningful page-1-to-page-2 risk while ignoring routine wobble.

​
Tracked live in Vortex IQ Nerve Centre

Ranking-Drop Alert is one of hundreds of KPI pulses Vortex IQ tracks across Google Search Console and 70+ other ecommerce connectors. Nerve Centre runs the detection layer; Vortex Mind investigates the cause when something moves; Ask Viq lets you interrogate any number in plain English. Start for free or book a demo to see this metric running on your own data.