AI citation share · Intent over position
Holding 100% of the AI citations for a question three million people ask
When somebody asks Bing Copilot when the Velankanni flag hoisting is in 2026, every citation in the answer points at this site. It also earns 42,901 Google clicks at a 3.2% click rate from an average position of 9.1 — the clearest evidence in the portfolio that intent quality, not position, decides what a ranking is actually worth.
100%
Copilot citation share on “velankanni flag hoisting date 2026”
16.3K
Bing Copilot citations in 6 months
86
Distinct pages cited by Copilot
3.2%
Google CTR, at an average position of 9.1
42,901
Organic clicks from 1,342,795 impressions
Property
VelankanniShrine.in · owned and operated
Type
Independent pilgrim guide, no official affiliation
Audience event
3M+ pilgrims at the annual feast, 29 Aug – 8 Sep
Window
16 months, ending 12 August 2026
Site size
662 pages ranking in Google
Paid promotion
None
The second finding
The weakest positions in the portfolio, and the best click rate by 3×
This property ranks worse than the other two on every positional measure, and outperforms both on the only measure that pays.
Where 1,000 the top 1,000 exportable queries actually sit
Only 2 queries out of 1,000 average position #1, and 46 average the top 3. By the standards of a rankings report this property is underperforming badly — while converting at 3.2%, three times the rate of my city-guide property. Search Console caps its export at 1,000 rows, so this is the distribution across the top 1,000 queries by clicks.
46
Top 3 · 4.6% of the queries measured
749
Page one · 74.9% of the queries measured
Source: Google Search Console, velankannishrine.in, Search type: Web, 16 months ending 12 August 2026. Band thresholds are ours: #1 is an average below 1.5, since Search Console reports position as a decimal average rather than an integer rank.
The explanation is in the query strings. “Velankanni shrine room booking”. “Shrine rooms booking office”. “Mass times”. These are people with a task to finish, and somebody trying to reserve a room for a pilgrimage will click past three results to find the page that actually explains the process.
So the booking-office page converts at 9.27% from an average position of 7.34, and one query — “velankanni shrine room booking” — converts at 19.28% from position 7.66.
Which is why chasing position 1 here would have been the wrong investment. The money was in making the page that ranks seventh unmistakably the one that completes the task.
Why the pages compound
One trip, seven searches, and the site owns the whole sequence
One pilgrimage, seven searches, one site
This is the order a visitor actually asks things in, with the real query and the page that answers it. Content built along this chain compounds, because answering question three is how you get asked question four.
Is it worth going, and when
“best time to visit velankanni”
/best-time-to-visit-velankanni/
How far is it
“chennai to velankanni distance”
/chennai-to-velankanni-train/
How do I get there
“chennai to velankanni train”
/chennai-to-velankanni-train/
What time does it leave
“chennai to velankanni train timings”
/chennai-to-velankanni-train/
When is the feast
“velankanni flag hoisting date 2026”
/velankanni-annual-feast-2026/
What time is mass
“velankanni church mass timings”
/velankanni-church-mass-timings/
Where do I sleep
“velankanni shrine room booking”
/explore/shrine-rooms-booking-office-velankanni/
The constraints
A spike of three million people, and no authority to claim
A guide to somebody else's shrine has to be trustworthy without ever pretending to speak for it.
No official standing, and a sensitive subject
The site has no affiliation with the basilica or the diocese, and says so plainly on every relevant page. It is a guide for pilgrims, not a religious authority — which means it has to be accurate enough to be trusted without ever claiming to speak for the shrine.
The demand is a spike, not a stream
Three million people attend the annual feast between 29 August and 8 September. The date, the flag hoisting, the mass schedule and every room in town become urgent questions for a few weeks, then go quiet. Get the page ready late and the year is gone.
The audience is multilingual and abroad
Tamil, English and Malayalam speakers across India, the Gulf, Singapore, Malaysia and the UK. 1,724 clicks arrived through Google's machine translation, which only happens if the page is clean enough to translate well.
What was actually built
Six decisions, and what each one bought
Own the date questions absolutely
Feast dates, flag hoisting, novena start — each one gets its own page with the year in the URL and the answer in the first line. These are the queries where being second is worth nothing, and where the citation share is now 84% and 100%.
Build the journey, not the topic
Distance, then trains, then timings, then mass schedule, then rooms. The pages follow the order a pilgrim actually asks things in, and each links to the next question rather than to a related category.
A page per place, at scale
Every hotel, lodge, guest house and pilgrim quarter has its own page with real details. 57 of those individual listing pages have been cited by Copilot — a long tail no single overview page could have earned.
Optimise for the task, not the rank
The booking-office page averages position 7.34 and converts at 9.27%, because somebody trying to reserve a room clicks past three results to find the one that actually explains the process. Chasing position 1 here would have been the wrong investment.
Write it so a machine can translate it
Short sentences, plain structure, no idiom in the load-bearing facts. That is why the diaspora reads it through Google Translate at a 2.97% click rate instead of bouncing off mangled text.
Say what you are not
An explicit independence notice, and no claim to official status. On a religious subject that disclosure is what makes the rest of the page credible — to readers and to the models deciding whether to quote it.
Evidence · AI search
16.3K citations across 86 pages
Note the two totals below and why they differ — the dashboard headline and the per-page export do not agree, and the honest thing is to publish both.
Bing · 16.3K Copilot citations, 11 cited pages on an average day
Citation sources: Microsoft Copilots and Partners, 6-month window. The step up in July is the run-in to the annual feast. The Grounding Queries tab below the chart is where citation share lives.
Why two different totals appear on this page
The dashboard reports 16.3K citations for the window. The per-page CSV export sums to 14,583 across 86 pages — about 89% of it. Bing states the report is a sample of overall activity and that page-level attribution is incomplete, so the two will not reconcile. Both figures are published here rather than picking the flattering one.
Copilot citations per page
From the per-page export. The homepage and the Chennai-to-Velankanni train page are within 200 citations of each other — a single transport page carrying as much AI visibility as the entire site's front door.
14,583
AI answer citations across all 86 cited pages, of which the top 10 are charted
Source: Bing Webmaster Tools AI page citation report, exported 12 August 2026. Top 10 of 86 cited pages shown; the full export includes a long tail of individual accommodation listings.
Citations by query intent
Bing classifies each grounding query by intent. Planning and Utility together are 55% of all attributed citations — people working out logistics, not reading about the shrine.
3,516
citations attributed to the 30 labelled grounding queries
Source: Bing Webmaster Tools grounding-query export, grouped by Bing's own intent label. Bing notes these labels are AI-generated. Three queries in the export carry no intent label and are excluded from this grouping, which is why this total is lower than the site's citation total.
The long tail is the part worth copying. Of the 86 cited pages, 57 are individual hotel, lodge and guest-house listings, earning 1,473 citations between them — a handful each.
No single “hotels in Velankanni” overview page could have collected those. They exist because each property has its own URL with its own real details, which is what makes it a distinct entity a model can cite. That is the argument for a page per entity instead of a roundup, and it is directly transferable to integration pages, template libraries and comparison pages on a SaaS site.
Evidence · Search Console
42,901 clicks from 1,342,795 impressions
Search Console · 16-month performance
42.9K clicks, 1.34M impressions, 3.2% CTR, average position 9.1. All four metrics are switched on so the CTR and position lines can be read against each other.
Search Console · top queries are all transactional
Room booking at 1,390 clicks, mass times at 1,374, and “velankanni shrine room booking online” converting at 18.39%. Every one of these is somebody with a trip to organise, which is why the click rate survives a mediocre position.
Top queries by clicks · 16-month averages
| Query | Clicks | Impressions | CTR | Avg position |
|---|---|---|---|---|
| velankanni church room booking | 1,390 | 22,166 | 6.27% | 8.3 |
| annai velankanni shrine mass times | 1,374 | 11,051 | 12.43% | 4.26 |
| velankanni shrine room booking online | 1,001 | 5,443 | 18.39% | 5.05 |
| shrine rooms booking office | 944 | 5,000 | 18.88% | 3.98 |
| velankanni shrine room booking online free | 584 | 4,506 | 12.96% | 4.87 |
| annai velankanni shrine | 576 | 12,566 | 4.58% | 4.18 |
| velankanni church rooms | 575 | 4,974 | 11.56% | 7.43 |
| hotel paradise velankanni | 494 | 11,699 | 4.22% | 5.08 |
| velankanni shrine room booking | 468 | 2,427 | 19.28% | 7.66 |
Read the CTR column against the position column. Four of these convert in double digits from positions between 4 and 8.
Top pages by clicks · Google, 16 months
| Page | Clicks | Impressions | CTR | Avg position |
|---|---|---|---|---|
| /explore/shrine-rooms-booking-office-velankanni/ | 12,890 | 138,992 | 9.27% | 7.34 |
| /velankanni-church-mass-timings/ | 5,817 | 230,362 | 2.53% | 8.86 |
| / | 4,532 | 273,921 | 1.65% | 10.41 |
| /velankanni-to-trichy-government-bus-2/ | 1,971 | 52,347 | 3.77% | 8.53 |
| /explore/hotel-paradise-velankanni/ | 1,097 | 28,778 | 3.81% | 5.43 |
| /hotels-in-velankanni/ | 754 | 94,715 | 0.80% | 17.67 |
| /explore/st-peter-and-paul-pilgrims-quarters-velankanni/ | 731 | 8,666 | 8.44% | 6.62 |
Search Console · pages
The booking-office page carries 12,890 of the site's 42,901 clicks. The pages export returned 662 rows — under Search Console's 1,000-row cap, so unlike the queries figure this one is a real total.
The diaspora signal · translated results
| Search appearance | Clicks | Impr | CTR |
|---|---|---|---|
| Translated results | 1,724 | 58,073 | 2.97% |
Google served 1,724 clicks as machine-translated pages at a 2.97% click rate. That only happens when the prose is plain enough to survive translation — an argument for short sentences that has nothing to do with style.
Evidence · Bing
Positions 2 to 4 on the engine that feeds Copilot
Bing ranks this property far better than Google does, which is not incidental — it is the retrieval layer behind the citations above.
Bing · 3.7K clicks, 99.1K impressions
Over 24 months at 3.76% CTR. “velankanni church” sits at position 3.00, and “velankanni church rooms online booking” converts at 32.04% from position 2.2.
Bing · the booking page at 16% CTR
The homepage averages position 3.98 on Bing against 10.41 on Google. The booking-office page converts at 16.04% from position 2.49 — the same task-completion pattern, amplified by a better position.
Bing top keywords · 24 months
| Keyword | Impr | CTR | Pos |
|---|---|---|---|
| velankanni church | 9.1K | 4.90% | 3 |
| velankanni | 8.6K | 0.63% | 4.71 |
| velankanni matha | 1.2K | 1.80% | 3.18 |
| velankanni church rooms online booking | 309 | 32.04% | 2.2 |
Bing by device · the inverse of Google
| Device | Impr | Clicks | CTR |
|---|---|---|---|
| Desktop | 88,976 | 3,256 | 3.66% |
| Mobile | 9,729 | 458 | 4.71% |
Bing is 90% desktop here while Google is 86% mobile. Same site, same content, opposite audience — a reminder that “our users are on mobile” is a per-engine statement, not a fact about your product.
Who it reaches
86% mobile, and a Gulf diaspora that converts better than home
By device · Google
| Device | Clicks | Share | CTR |
|---|---|---|---|
| Mobile | 36,816 | 85.8% | 3.27% |
| Desktop | 5,938 | 13.8% | 2.84% |
| Tablet | 147 | 0.3% | 2.14% |
By country · Google
| Country | Clicks | CTR | Pos |
|---|---|---|---|
| India | 41,319 | 3.37% | 8.48 |
| United Arab Emirates | 429 | 4.70% | 9.05 |
| United Kingdom | 121 | 1.81% | 24.62 |
| Singapore | 119 | 2.63% | 8.97 |
| Malaysia | 106 | 1.63% | 8.9 |
| Kuwait | 67 | 3.35% | 9.5 |
The UAE converts at 4.70% against India's 3.37% — migrant workers planning a pilgrimage home, on a shorter decision window and with less patience for a vague page.
Coverage, stated honestly
What is proven here, and what is missing
This property has a gap the other two studies do not: no Google AI report exists for it. It is marked as missing rather than filled in.
| Surface | Status | Finding | Evidence |
|---|---|---|---|
| Google organic | Documented | 42,901 clicks from 1.34M impressions at a 3.2% CTR — three times the click rate of the other two properties, on the weakest average position of the three. | Search Console exports + screenshot |
| Google translated results | Documented | 1,724 clicks from 58,073 impressions served as machine-translated pages, which is the diaspora reading in its own language. | Search Console Search Appearance export |
| Bing organic | Documented | 3,714 clicks from 98,705 impressions at 3.76%, averaging positions 2.2 to 4.7 on head terms. One booking query converts at 32%. | Bing Webmaster Tools exports |
| Bing Copilot | Documented | 16.3K citations across 86 distinct pages, with 100% citation share on one event query and 9 of 33 grounding queries above 50%. | Bing AI Performance + grounding-query export |
| Google AI features | Not audited | Search Console does not expose a Generative AI report on this property, so unlike the other two studies there is no Google-side AI figure to publish. Its absence is stated rather than filled in from elsewhere. | Report unavailable |
| ChatGPT, Gemini, Perplexity | Not audited | No per-answer citation export exists for these surfaces, so they are not measured or claimed. | No export available |
Why a pilgrim guide is in a SaaS SEO portfolio
Three transfers, and one I now do on every audit
Date and version questions are winnable absolutely. The reason this site holds 100% citation share is that a feast date has exactly one right answer, stated specifically, on a page whose URL names the year. The SaaS equivalents are release notes, pricing changes, deprecation timelines and compliance deadlines — questions your customers ask an assistant, where being the definitive source is achievable this quarter.
A page per entity beats a roundup. Fifty-seven individual accommodation pages earned citations a single “hotels in Velankanni” page never would have. Your integrations, templates, connectors and comparison pages are the same shape: give each one a URL and real detail, or accept that none of them get cited.
Read alongside the Coimbatore Junction study, these two properties are a natural experiment: near-identical positions, click rates a hundredfold apart, and the difference is entirely the intent behind the query. That is the pair of data points that made me stop reporting a single site-wide CTR.
Who did this work
An owned property, published in full
Alston Antony
Owner & SEO lead
VelankanniShrine.in is owned and operated by Alston Antony. It is an independent pilgrim guide with no affiliation to the basilica or the Diocese of Thanjavur, and it states that on the site rather than leaving it ambiguous. On a religious subject that disclosure is what makes the rest of the page credible.
He has worked in search since 2010, holds an MSc in Computer Software Engineering (Distinction) from the University of Greenwich, and is a professional member of BCS, The Chartered Institute for IT. His day role is Senior Digital Marketing Manager at Brainstorm Force, whose software powers 7M+ sites.
How to verify every number on this page
- Google totals come from Search Console for velankannishrine.in, Search type: Web, the 16-month window ending 12 August 2026. Site totals use the Devices export, which is complete.
- Pages is a real total; queries is a cap. The pages export returned 662 rows, under Search Console's 1,000-row ceiling. The query export hit the cap at 1,000, so every position band is a share of that sample.
- Citation share is Bing's own figure from the grounding-query export, not a calculation of mine. I filtered the chart to queries with 25+ citations so a single citation at 100% could not be presented as dominance.
- Two AI totals are published because they disagree: the dashboard says 16.3K, the per-page export sums to 14,583. Bing states the report is a sample, so neither is a hard number and I did not pick the larger one and stop.
- Intent labels are Bing's, and AI-generated — its own interface says so. Two grounding queries carry no label and are excluded from the intent grouping rather than bucketed as “other”.
- No Google AI figure appears because Search Console exposes no Generative AI report on this property. That absence is in the coverage table.
- The site is live. Open velankannishrine.in and ask Copilot when the 2026 flag hoisting is.
Case study FAQ