The long tail advantage is 2.27x, and narrower inside a position band

Chase the long tail. It is long-standing advice in search, and we believed a strong version of it: our own notes recorded a head term at 0.06 percent click-through against a specific variant of the same subject at 7.50 percent, a gap of roughly 125 times.

This is the check of that claim. On one cluster in one window the rarer queries did have higher combined click-through, by far less than that.

The whole cluster, by query volume

One cluster, one window, 3,328 queries. Band them by how many impressions each query drew and the pattern matches the advice.

Read end to end that is 5.3x, and taken as it stands it reads as an argument for chasing the tail.

That 5.3x is not the number this finding tests. It compares the single query above 100,000 impressions against everything under 10. The table split by position band further down uses wider cohorts, 1,000 or more against under 100, so setting 5.3x beside it would change the cohorts and the banding at once. On the cohorts used below, the ratio across all positions is 2.27x: a combined 0.1103 percent for queries with 1,000 or more impressions against 0.2499 percent for those under 100. That is the figure the within-band ratios are set beside.

The position column

Read the average position column. It drops row by row as volume drops, from 8.96 for the largest query to 5.90 for queries under 10 impressions. On the two cohorts this finding uses, the head averages 8.75 and the tail 7.67, so the rarer queries also average about one position better. Volume and average position both differ between the cohorts, and the table is split by only one of them.

Comparing queries within one position band

Cross-band the same queries by average position and by volume together. Site and window are the same throughout; the cohorts may still differ in query intent and in the pages receiving impressions. Finding 08 makes the same kind of within-band comparison for pages, in the position 5 to 10 band.

The same cohorts that give 2.27x across the cluster give 1.57x at average positions 5 to 8 and 1.23x at 8 to 11. Both within-band figures are smaller than the cluster-wide one, and far smaller than the 5.3x read end to end from the first table.

Mid to tail: nearly equal in one band, lower in the other

Look at the step from mid to tail rather than from head to tail, at average positions 5 to 8 and then 8 to 11. The observed ratios are approximately 1.00x and 0.86x.

At 5 to 8 the mid and tail cells carry 47 and 31 clicks; at 8 to 11, 84 and 33. These small counts do not establish equal underlying rates or a lower underlying rate for the tail.

What can be said is narrower. In both bands, queries with 100 to under 1,000 impressions already had higher combined click-through than the head cohort, and the tail cohort's rate was nearly equal to theirs in one band and lower in the other.

What is left standing

A difference, and a far smaller one than our notes recorded. Inside the two average-position bands, queries with 100 to under 1,000 impressions had 1.57 and 1.43 times the head cohort's combined click-through, and queries under 100 had 1.57 and 1.23 times. That describes these cohorts in this window, not what targeting more specific queries would do for a page.

The other thing in the data, and it is an observation rather than a mechanism: the rarer queries also average better positions. Head and tail average 8.75 and 7.67 across the cluster. Inside the 5 to 8 band they still differ, 7.42 against 6.87; inside 8 to 11 they are nearly equal, 9.06 against 9.10. Whether average position accounts for any of the click-through gap, or both follow from something else about specific queries, is not something this comparison can separate. Doing so needs an adjustment model fitted to the position distributions, and none was fitted here.

The 125x in our own notes

Our own record held a head term at 0.06 percent against a specific variant at 7.50 percent, about 125 times. On this window the best query above 100 impressions reaches 2.83 percent against a head term at 0.105 percent, which is 27 times, and it rests on three clicks.

The 125x does not reproduce in this window. It came from a different window, it was never split by position, and a rate computed from a handful of events is the thing this site exists to refuse. It should not be requoted, including by us.

The limit that cuts against this finding

This export contains 636,317 impressions against 810,421 counted across the cluster's canonical URLs for the same window, so roughly 21 percent of the impressions are not returned. It returns 820 of the cluster's 1,793 clicks, 46 percent.

What the missing impressions and clicks consist of, we do not know. Search Console anonymises rare queries, and it also documents additional loss when results are grouped by query rather than reported in aggregate. Calling the gap the rarest queries by definition would assume a composition we never measured, and the one most convenient to the argument.

The honest position: the deep tail measured here is the tail the instrument returned, and about a fifth of the cluster's impressions and more than half of its clicks sit outside it, for reasons we cannot decompose. That is a limit on the conclusion in both directions.

How to check your own

Pull queries for one group of related pages, for example by a shared URL term, and keep in mind that the volume cohorts can still differ in query intent and in the pages receiving impressions. Band them by impressions and look at the average position column before anything else. If average position moves with volume, a comparison across all positions mixes the two.

Then cross-band by average position and volume together, sum clicks and impressions inside each cell, and read the step from mid to tail as well as from head to tail. On this cluster the rates at that step were nearly equal in one band and lower for the tail in the other.

Full method

Google Search Console searchAnalytics, dimension query. Queries are restricted to pages whose URLs contain one shared path term for bank branch codes, within one site and reporting window. The cohorts may still differ in query intent and the pages receiving impressions. The headline ratios compare the same two volume cohorts throughout, 1,000+ impressions against under 100; a middle cohort of 100 to under 1,000 is shown beside them. Rates are combined: a cohort's clicks summed over its impressions summed. Bands use each query's average position, from 5 up to 8 and from 8 up to 11, and a cohort's average position is weighted by impressions. Figures are reported descriptively: an overall ratio beside two within-band ratios, with no adjustment model fitted and therefore no explained fraction claimed.