16 months of Search Console data · April 2025 to July 2026 · Measured clicks
The measured data confirms the diagnosis, and corrects the numbers.
Clicks are down 71% year on year with no recovery. Almost all of it is the blog library. The campus and locality pages that drive admissions held, and brand demand never fell.
In July we analysed the decline using a third-party estimate, because Search Console access did not exist yet. The exports you shared let us check that work against Google's own click data. The main finding held. Five of the numbers around it did not.
What the July analysis said
Verdict
What the measured data shows
Traffic recovered about half of the decline between May and July 2026
Corrected
It did not. Clicks were flat at about 546 a day in April 2026 and 531 a day in July 2026. The recovery appears only in the estimate, which prices ranking position rather than clicks.
Traffic eroded gradually through 2025
Corrected
Clicks held broadly steady from April to November 2025. The whole decline happened in two steps between November 2025 and April 2026, and has been flat since.
The blog library is about 70% of the loss, campus pages about 23%
Corrected
The concentration is far sharper. Blog traffic fell 90% and accounts for 93% of the loss. Campus pages fell only 7%.
About 27% of the decline is branded search, probably falling brand demand
Corrected
Branded search is 3% of the loss. On a like-for-like set of 474 branded searches, the site was shown 1.7% more often than a year ago. Demand is intact; the ranking for it slipped slightly.
Around 80 pages appear to have been dropped from Google
Corrected
Six pages were deliberately removed and now return a permanent removal response. They alone carry 14% of the loss. Of 238 pages checked, 232 are live and serving normally.
Traffic fell about 61% from the January 2025 peak
Refined
Measured clicks are down 71% year on year. The decline is deeper than the estimate suggested. The January 2025 peak itself sits before the records supplied, so it cannot be checked here.
AI answers are intercepting clicks on informational searches
Refined
Partly. But searches where the site lost visibility outweigh searches where it held its place and lost clicks by about five to one. This is mainly a visibility problem.
Backlinks and seasonality are not the cause
Confirmed
Nothing in the measured data contradicts this. Comparing the same four months a year apart removes the admissions season from the picture, and the decline survives that test.
The estimate tracked reality closely for fourteen months, then diverged in June and July 2026. It priced a genuine ranking improvement on very general searches that deliver almost no clicks, and read it as a recovery.
What happened
Clicks fell 71%, in two steps, and have not recovered.
Traffic was broadly steady from April to November 2025. It then fell in two distinct steps, between November 2025 and January 2026, and again between March and April 2026. Since April it has been flat at roughly 530 clicks a day.
-71%Clicks a day, April to July 2025 compared with the same four months in 2026
93%Share of the entire loss that came from the blog library alone
-7%Campus and locality pages, the section that drives admissions
FlatBrand demand. People search for EuroSchool as often as they did a year ago
Daily clicks by type of search. The blue band at the bottom is branded search, which barely moves across the whole period.
Why two steps matters
A gradual slide would point at slow content decay with no single cause to find. Two sharp steps five months apart point at specific, datable events. That is a much smaller area to search.
Where it went
The blog library is 93% of the loss. The admissions pages are 2%.
Splitting every page on the site into its section makes the shape unmistakable. One section fell away. Almost everything else either held or grew.
Clicks a day gained or lost by section, April to July 2025 compared with the same months in 2026.
Section
Clicks a day, before
Clicks a day, now
Change
Share of the loss
Blog library
1,328
134
-1,194 (-90%)
93%
Syllabus and study guides
39
11
-28 (-72%)
2%
Campus pages
280
259
-21 (-7%)
2%
Calculators and converters
20
1
-19 (-94%)
2%
City pages
29
11
-19 (-64%)
2%
Homepage
58
49
-9 (-16%)
1%
Why EuroSchool
17
26
+9 (+55%)
grew
Careers
44
47
+3 (+7%)
grew
Fee cards and disclosures
4
7
+2 (+55%)
grew
In March 2026 the campus and city pages overtook the entire blog library, not by growing but by holding steady while the blogs fell away.
The ten blog posts behind most of the decline
These ten alone account for about a quarter of the site's total loss. Four of them no longer exist: they were deleted and now tell Google to remove them permanently.
Campus pages declined gently and evenly, between 0% and 35%. An even spread like this is a mild headwind across the site, not a problem with any one campus. Gunjur, Hennur and Manjri are new and already earning searches from a standing start.
Campus
City
Clicks a day, before
Clicks a day, now
Change
EuroSchool HSR Layout (ICSE)
Bengaluru
15.0
17.9
+19%
EuroSchool Wakad (CBSE)
Pune
18.0
17.9
0%
EuroSchool Yelahanka North
Bengaluru
15.5
15.1
-2%
EuroSchool Wakad (ICSE)
Pune
17.8
17.1
-4%
EuroSchool Whitefield
Bengaluru
31.6
28.6
-10%
EuroSchool Electronic City
Bengaluru
15.5
13.9
-10%
EuroSchool Airoli (ICSE)
Navi Mumbai
18.4
16.1
-12%
EuroSchool Undri (ICSE)
Pune
14.5
12.0
-17%
EuroSchool Ghodbunder Road
Thane
16.8
13.8
-18%
EuroSchool Chimney Hills
Bengaluru
11.3
9.0
-20%
EuroSchool Bannerghatta
Bengaluru
15.3
11.9
-22%
EuroSchool Dombivli
Thane
14.6
11.3
-23%
EuroSchool Upper Thane
Thane
5.9
4.3
-26%
EuroSchool Balkum
Thane
11.2
8.1
-27%
EuroSchool Hitec City, Gachibowli
Hyderabad
16.8
11.8
-30%
EuroSchool Metro Campus, Kukatpally
Hyderabad
15.7
10.7
-32%
EuroSchool Kharadi
Pune
23.8
15.4
-35%
EuroSchool Hennur
Bengaluru
0.0
9.0
New
EuroSchool Manjri
Pune
0.0
7.7
New
EuroSchool Gunjur
Bengaluru
0.0
5.3
New
Campus and locality searches
Ranking better. Clicked far less.
You asked which campus keywords lost their rankings. Almost none did. Taking the 79 non-branded locality searches that appear in both years, so the comparison is like for like, EuroSchool now ranks more than seven places higher and is shown more than twice as often. It still gets a third fewer clicks, because the click-through rate on those same searches fell by nearly three quarters.
13.4 to 6.0Average position on locality school searches, a gain of more than seven places
+132%How often those searches now show EuroSchool
-73%Click-through rate on the very same searches
-36%Clicks, despite everything above moving the right way
The same set of searches on all three panels. Visibility and ranking have improved steadily since late 2025. The click-through rate has not moved since it fell in June 2025.
The month it changed
Click-through on these searches went from 9.2% in May 2025 to 1.4% in June 2025, and has not been above 2% in any of the fourteen months since. Branded click-through held steady through that same month, at 10% to 13%, so this is not a measurement problem or a tracking change. Something altered what the results page looks like for non-branded searches. A map pack or an AI answer above the results would both produce exactly this pattern, though Search Console does not report either, so we cannot confirm which from this data alone.
The campus searches that lost the most clicks
Read the position column. Most of these rank better now than they did a year ago. "best school in wakad pune" went from tenth place to first and still lost 84 clicks. Only five searches in the whole set genuinely lost rank or disappeared, and they are marked below.
Search
Clicks before
Clicks now
Position before
Position now
What happened
best schools in bangalore
260
17
20.8
9.6
Ranks better
best cbse school in whitefield bangalore
220
0
13.6
gone
Disappeared
cbse schools in yelahanka
182
10
13.3
5.1
Ranks better
best schools in kukatpally hyderabad
181
5
9.0
3.3
Ranks better
best schools in kharadi pune
181
6
3.1
6.4
Lost rank
schools in undri pune
151
14
10.1
6.2
Ranks better
cbse schools in hitech city
136
0
4.4
gone
Disappeared
best cbse school in kharadi
124
0
5.4
gone
Disappeared
cbse school in metro campus kukatpally
90
0
1.0
gone
Disappeared
best school in wakad pune
86
2
10.5
1.0
Ranks better
cbse schools in kharadi pune
72
3
4.9
6.0
Held rank
best icse schools in pune
70
10
6.1
3.9
Ranks better
And the ones that grew
City-level searches are where the ranking gains did convert. EuroSchool climbed from the third page to the first on several of them.
Search
Clicks before
Clicks now
Position before
Position now
schools in bangalore
13
143
26.3
7.4
icse schools in pune
30
77
6.8
2.1
bangalore schools
0
51
new
1.8
schools in pune
3
50
33.8
5.1
cbse schools in bangalore
0
49
new
12.3
icse schools in bangalore
4
47
13.6
6.5
cbse schools in electronic city
0
41
new
3.7
cbse schools in whitefield
0
31
new
5.2
Area by area
Six areas lost clicks and six gained. Nine of the twelve rank better than they did a year ago, including four of the six that lost clicks. The last column is the change measured on the campus pages themselves, which is a completely separate measurement: ten of the twelve agree, which is a good sign that the picture is real.
Clicks gained or lost per area, with each area's ranking change alongside.
Area
Clicks change
Position before
Position now
Campus page traffic
Kharadi
-313
4.4
3.0
-53%
Kukatpally
-186
6.8
2.9
-38%
Whitefield
-167
9.8
10.0
-53%
Yelahanka
-151
12.4
4.8
+22%
Undri
-104
6.0
6.3
-12%
Gachibowli and Hitec City
-85
6.1
4.6
+1%
HSR Layout
+40
7.5
3.2
+287%
Wakad
+55
6.7
5.7
+60%
Airoli
+64
4.2
4.7
+60%
Bannerghatta
+103
5.4
6.3
+76%
Dombivli
+128
6.8
3.7
+39%
Electronic City
+152
13.7
3.5
+323%
The one part that did break
The city index pages are the exception. Non-branded traffic to the pages that list all schools in a city fell 93%, and Bengaluru alone went from 360 clicks to 5. Those pages sit between a parent searching and the campus they actually want, and they have stopped earning their place. Everything else in this section is about campus pages, which grew their non-branded traffic 17% over the same period.
How this was measured
Search Console reports searches and pages separately, so a search cannot be tied directly to the page it landed on. Searches were matched to an area by the words in them, then checked against what the campus pages themselves recorded. Ten of the twelve areas agree between the two, which is why we are confident in the pattern. Two do not: Yelahanka, where the searches fell but the page grew, and Gachibowli and Hitec City, which share one page. The headline uses only searches present in both years, so a change in which searches happen to be reported cannot affect it.
Why it happened
Three causes, in the order they matter.
These do not overlap. Each number is a separate slice of the same decline, and together they account for essentially all of it.
Primary93% of the loss
The blog library stopped ranking
Around 1,500 informational posts, mostly general-knowledge lists and school-topic explainers, lost 90% of their clicks. Of the declining pages we checked, 232 of 238 are live and serving normally, so this is Google choosing not to rank them rather than anything broken. Six of the largest were deliberately deleted and now tell Google to remove them permanently.
Not what it looked like3% of the loss
Branded search held, but the ranking for it slipped
On a like-for-like set of 474 branded searches, the site was shown 1.7% more often than a year ago, so demand for the EuroSchool name has not fallen. What changed is the average position, from 1.98 to 2.28. At the very top of the results a third of a place is worth a lot of clicks. Something is now appearing above EuroSchool on its own name plus locality searches.
Not admissions traffic27% of the loss
Overseas visitors and a Google feature withdrawal
Searches from outside India fell 86% and make up more than a quarter of the headline decline. A large part of that is Google withdrawing its automatically translated results, which fell from about 6,200 clicks a month to under 300. This traffic was never going to enrol a child in Bengaluru or Pune, so it should be excluded from any recovery target.
The top chart is how often Google shows the site. The bottom is how often people click when it is shown. The first collapsed; the second did not. That means the site is being found less, not passed over more.
What this is not
It is not a penalty on the whole site, and it is not the admissions pages. Campus and locality pages fell only 7% while three new campuses started earning traffic from scratch. Careers, Why EuroSchool and the fee-card downloads all grew. It is also not primarily AI answers taking the clicks: searches where the site lost visibility outweigh searches where it held its place and lost clicks by about five to one.
That balance holds for the site as a whole, because the blog library is large enough to set the totals. For campus and locality searches the reverse is true: visibility grew and it was the click-through that collapsed. The section above sets that out.
A caution on reading this: the top-three share rising from 42% to 62% is not an improvement. The searches that used to sit further down the first page have left the picture, so the ones still visible are the strongest ones. The site ranks well for what remains, and appears for far less.
Way forward
Two questions for you, and one thing we can start today.
The two largest findings both point at changes made to the site rather than anything Google did. Before we plan around them, we need to know whether they were intended.
Needs an answer from your side
Six pages were deleted
Six blog posts now return a permanent removal signal to Google, and they account for 14% of the entire decline. One of them was the single biggest page on the site. We need to know whether that removal was intended. If it was not, it can be reversed.
Needs an answer from your side
Publish dates have been reset
About a quarter of the blog posts we sampled now declare a 2026 publish date although they were already ranking in 2025, and almost all of them still carry a lead image from 2023 or 2024. Google asks that dates reflect real updates. We need to find what produced this before deciding how to correct it.
We can start now
Protect and grow the pages that held
Campus and locality pages are the part of the site closest to admissions and they came through the decline nearly intact. Three new campuses are already picking up searches. This is where added depth pays back fastest, and it is the locality-first direction we already agreed.
The full plan, in priority order
Priority
Action
Why it matters
Must-fix
Confirm whether the six deleted pages were meant to go, and restore them if not
Six URLs carry 14% of the total decline, including the largest single page on the site. A permanent removal response is the strongest instruction you can give Google.
Must-fix
Find what reset the publish dates on the blog library and correct it
Roughly a quarter of sampled posts claim a 2026 publish date while their images and their traffic history say 2023 or 2024. Google's guidance is that dates should reflect genuine updates.
Must-fix
Win the click on locality searches, not the ranking: a complete and actively managed Google Business Profile for every campus, steady review volume, presence in the map results, and how each result itself reads in the listing
Ranking is not the constraint here. EuroSchool already climbed from position 13.4 to 6.0 on these searches and is shown twice as often, and the clicks still fell. Click-through is where the ceiling now sits.
Must-fix
Identify who now outranks EuroSchool on branded name plus locality searches, and win the position back
This is the branded case, and it is a genuine ranking loss: position slipped 1.98 to 2.28 while demand held. Highest-intent traffic on the site, and every place lost is recoverable.
Must-fix
Run a full crawl and check indexing coverage, sitemap, robots rules and canonical tags
The site is shown 63% less often than a year ago. Being found again comes before being clicked again.
Should-fix
Decide the future of the blog library: merge near-duplicates into fewer genuinely useful guides, put named authors and real school expertise behind them, refresh what earns it and retire the rest
The pages are live and simply not being ranked. That is a quality judgement, and the honest answer is fewer, better pages rather than more of the same.
Should-fix
Correct the campus pages that advertise 2026-27 in the title while their date and photograph still say 2025
Eighteen of the twenty-five campus pages checked have this mismatch. These are the admissions pages, and the fix is quick.
Should-fix
Make the surviving content easy to quote: structured data, a direct answer near the top of each page, clean heading order, and access for the AI crawlers
Answer engines are a growing share of how parents research schools. This positions the content that survived rather than fighting the trend.
Nice-to-have
Treat the overseas and translated-results decline as out of scope
More than a quarter of the headline number, with no admissions value, and partly caused by Google retiring a feature.
What a realistic recovery looks like
Not all of the 71% is winnable. More than a quarter of it was overseas traffic with no admissions value, and some of the informational searches now get answered inside Google. The traffic worth chasing is the Indian, parent-intent share: the campus and locality pages that already hold, the branded searches where demand never fell, and a smaller, better blog library that earns its place.
Context
How to read these numbers.
01
These are measured clicks
Every number here comes from Google Search Console, which counts actual clicks. The July analysis used a third-party estimate because no Search Console access existed at the time.
02
Compared like for like
April to July 2025 against the same four months in 2026, measured per day. That removes the admissions season and the fact that months differ in length.
03
The headline overstates the problem
More than a quarter of the decline is overseas traffic that would never lead to an admission. India on its own is down 67%, and the campus pages are down 7%.
04
One limit worth knowing
The records supplied begin in April 2025, so the January 2025 peak referenced in the earlier analysis sits before this data and cannot be checked against it.
Next steps
What we need from you.
Two answers unlock the two biggest items on the plan. Everything else we can begin without waiting.
Confirm the deletions and the dates
Whether the six removed blog posts were meant to go, and who or what reset the publish dates across the blog library. Between them these two questions cover the largest recoverable part of the decline.
Search Console access when you can
These exports carry us a long way. Direct access would add indexing coverage, any manual actions, and live weekly tracking so we can see a recovery as it happens rather than a month later.
