Skip to the indexed answers
Digital Marketing Index

usdigitalmarketingindex.com SEO

How can you identify pages whose organic traffic depends on a single query?

By calculating, for each page, the share of its search clicks that come from its top query, and listing the pages where that share is high. Search Console provides the data: filter the performance report to one page, open the queries table, and divide the clicks from the leading query by the page’s total clicks for the same period. A page that gets 85 of every 100 clicks from one query has its fortunes tied to that query. If demand for it falls, if a competitor takes the top position, or if a feature on the results page starts answering it directly, the page loses most of its traffic at once. The calculation takes minutes per page and can be run across a site from an export. Its result is a list of pages to watch and, where the page can honestly serve more questions, to broaden.

Query concentration calculation

The measure is simple.

Concentration = clicks from the page's top query / total clicks to the page

Procedure for one page:

  1. In Search Console’s performance report, add a page filter for the exact URL.
  2. Set a period long enough to smooth daily variation, such as three months.
  3. Open the queries tab and sort by clicks.
  4. Note the top query’s clicks and the total clicks shown for the page.
  5. Divide.

For many pages, export page and query data together through the API or a connected export, then compute the top-query share for every page.

A page-query concentration example, using hypothetical observations for one period:

Page Total clicks Top query Top query clicks Concentration Second query clicks Queries with at least 10 clicks
/guides/boiler-pressure/ 4,000 boiler pressure too high 3,400 85% 180 6
/boiler-repair/ 2,500 boiler repair 700 28% 450 31
/guides/bleed-a-radiator/ 1,800 how to bleed a radiator 1,170 65% 240 12
/power-flush/ 600 power flush cost 150 25% 120 18
/guides/boiler-error-e119/ 900 e119 boiler error 855 95% 20 2

Reading: the first and last pages are highly concentrated. The second and fourth draw on many queries.

Three refinements make the figure more informative.

  • Top three share. Add the top three queries’ clicks. A page at 40% top-query share and 90% top-three share is still narrow.
  • Query count. The number of queries bringing a meaningful number of clicks shows breadth directly.
  • Group close variants. “boiler pressure too high” and “boiler pressure high” are one need. Decide whether to treat variants as one query, and apply the decision consistently.

Limits of the data, to state with the results:

Limit Effect
Anonymized queries Search Console omits rarely searched queries from the query table. The clicks listed by query can therefore sum to less than the page's total, so compute concentration against the page's total clicks, and record the unlisted share
Row limits The interface shows a limited number of rows; long-tail queries may be cut off
Low totals A page with few clicks gives an unstable percentage
Period A short window exaggerates concentration; use the same window for every page
Clicks, not impressions A page may be shown for many queries and clicked for one

Because of the first limit, a page can look more concentrated than it is among visible queries and less concentrated than it is in total. Report three numbers: top-query clicks, total clicks, and clicks not attributed to any listed query.

Exposure to demand changes

Concentration matters in proportion to what depends on the page.

Factor Higher exposure when
The page's share of site traffic The page is among the site's largest
The page's role in conversions It brings leads or sales
The query's stability Demand is seasonal, tied to a news event, or to a product with a limited life
Competition for the query Strong competitors rank just below
Results page for the query The answer is short enough to be given on the results page itself
Position The page holds the top position, so any movement is a loss
Type of query A single error code, a single product model, or a single date-bound topic

An exposure view for the example:

Page Concentration Share of site clicks Query stability Exposure
/guides/boiler-pressure/ 85% Large Seasonal, recurring High
/guides/boiler-error-e119/ 95% Small Tied to one product model High for the page, low for the site
/guides/bleed-a-radiator/ 65% Moderate Seasonal, recurring Moderate
/boiler-repair/ 28% Large Steady Low

Concentration is not a fault. A page written to answer one question should get its traffic from that question. The purpose of the list is to know where a single change outside the site’s control would be felt, and how much.

Adjacent query opportunities

For concentrated pages that matter, look for queries the page could also serve.

Sources:

Source What to look for
The page's own query list, sorted by impressions Queries with many impressions and few clicks, where the page appears low
Queries containing the top query plus another word Refinements: causes, costs, steps, models, comparisons
Related questions shown on the results page What searchers ask next; record observations with dates
Customer questions from support records Real follow-up needs
Other pages on the site Whether another page already serves the adjacent query

Then decide per query.

Situation Action
The adjacent question belongs to the same need and the page does not answer it Add a section to the page
The adjacent question is a different need A separate page, linked from this one
Another page on the site already answers it Link to it; do not duplicate
The page already appears for the query at a low position Improve the relevant section and its heading
The query is unrelated Leave it

Broadening has a limit. Adding sections for loosely related queries dilutes a page that works. The test is whether a reader who arrived for the main query would find the addition a natural part of the answer.

For the boiler pressure page in the example, adjacent needs within the same visit might be low pressure, how to repressurize, and when to call an engineer. Whether those belong on the page or on pages of their own depends on how much each needs to say. Any instructions involving the appliance itself should come from the manufacturer’s documentation or a qualified source.

Monitoring risk

Track the concentrated pages that matter.

Measure Frequency Alert when
Concentration, by page Quarterly Rises substantially
Position for the top query Weekly Drops from its usual range
Clicks and impressions for the top query Weekly Clicks fall while impressions hold, or both fall outside seasonal expectation
Results page for the top query Monthly observation, recorded with date and device A new feature appears above the result
Number of queries with meaningful clicks Quarterly Falls
Page's share of site clicks Quarterly Rises, meaning the site as a whole is becoming more dependent

A site-level view is worth adding: the share of all organic clicks that comes from the ten largest page-and-query pairs. If that share is high, the site’s traffic rests on a few results.

A register to maintain:

Page Top query Concentration Why it matters Mitigation Owner Last reviewed
[page] [query] [%] [traffic, leads] [sections added, related pages built, none needed] [name] [date]

When a concentrated page does lose its query, the register shortens the diagnosis. The cause is already narrowed to one query, and the question becomes what changed for it: demand, position, or the results page.

This analysis works page by page on traffic the site already has. Judging which new queries a site can realistically win, and how to divide effort among them, are planning questions treated in the guides to whether a new website can realistically compete for a target keyword and how a limited SEO budget should be allocated between new content and technical fixes.

The method in order: compute each page’s top-query share over a fixed period, note the limits of query data, rank by what depends on the page, broaden where the reader’s need allows, and monitor the rest.

Link trail

Reader comments on this answer

Add your comment to this answer

Your email address will not be published. Required fields are marked *