49%
Of consumers used AI for savings or investment decisions in the last six months
Source: EY Global AI Sentiment Survey, 18,152 respondents
20%
Express any trust in AI within financial services, the lowest of any sector
Source: YouGov
30-40%
Visibility gain from citing sources, adding quotations and statistics
Source: Aggarwal et al., ACM KDD 2024
91%
Of week-on-week citation changes in finance are declines
Source: BrightEdge, most volatile category tracked
About GEO for Financial Services
The engines answer questions about money. They refuse to give advice about it. Everything follows from that line.
The Trust Paradox Nobody Is Exploiting
Four independent studies say the same thing. EY found 49% of consumers used AI for savings and investment decisions in the past six months. YouGov found only 20% express any trust in AI within financial services, the lowest reading across every sector they measure. TD Bank found 78% use AI tools while just 18% would trust it with a financial recommendation.
Read those together. People consult the machine, then look for someone credible to confirm what it said.
That confirmation is what a licensed institution can supply and a generative engine cannot. Not as a marketing claim. As a structural fact about who carries regulatory accountability.
Which Means the Goal Is Citation, Not Recommendation
Ask ChatGPT whether you should take a particular loan and it will decline, hedge, and suggest consulting a licensed advisor. Ask it how effective interest is calculated in Indonesia and it will answer, with citations.
The difference is not the topic. It is the difference between a question and a request for advice. Financial institutions can only occupy the first, and that is where fintech SEO and GEO stop being separate disciplines. A GEO company that treats them separately has already misunderstood the category.
The Engines Answer Questions. They Refuse Advice.
Same topic. Same product. Same user. One prompt gets a cited answer, the other gets a disclaimer. Understanding why that line exists is the entire strategy.
Apakah saya sebaiknya ambil KTA dari bank ini? Bunga 1,2% flat per bulan.
I cannot advise on whether to take a specific loan. That decision depends on your income, existing obligations and financial circumstances, which I do not have. Consider consulting a licensed financial advisor or the institution directly.
No citation. No brand mentioned. Nothing to optimise for.
Bagaimana cara menghitung bunga efektif dari KTA bunga flat 1,2% per bulan?
Flat interest is calculated on the original principal throughout the tenor, while effective interest accounts for the declining balance. A 1.2% flat monthly rate over 24 months is roughly equivalent to 22% effective annually, though the exact figure depends on repayment structure.
Cited, with sources. This is the space institutions can occupy.
OJK Edukasi
Bank explainer
Aggregator guide
Illustrative simulation of engine behaviour on advice versus information prompts. Figures are approximate.
Four Engines, Four Different Ideas of Who to Trust
Ask each one the same financial question and watch where they go looking. A strategy built for one engine tells you almost nothing about the others.
ChatGPT
Financial data aggregators
Established consumer finance publishers
Wikipedia, increasingly
Institution sites, when structured
AI Overviews
Consumer education content
Regulator and government sources
Comparison aggregators
Withdraws entirely on real-time data
Gemini
Government and institutional sources
Knowledge Graph entities
Brands with verified Knowledge Panels
Cites Reddit in 0.1% of responses
Perplexity
Review aggregators and forums
Recency weighted heavily
Requires corroboration across sources
Numbered citations, most measurable
Perplexity holds roughly 11 to 21 percent of Indonesian AI chatbot share against a global average nearer 2 to 3 percent, because Telkomsel bundled Perplexity Pro into its data plans from 28 May 2025. A GEO strategy copied from a Western agency underweights it badly.
In September 2025, a Configuration Change Erased a Channel
Reddit's citation share inside ChatGPT collapsed. Five trackers measured it. Five reported different numbers. All five were correct, and that is the part worth understanding.
Spotlight
14.29% to 0.21%
Share of all cited sources, across 3 million citations tracked between August and October 2025
SEMrush
~60% to ~10%
Share of prompt responses containing a Reddit citation, 230,000 prompt tracking study
Agency via RBC
29.2% to 5.3%
Methodology undisclosed. Reported to investors, moved the stock price
PromptWatch
9.7% to 2%
Share of ChatGPT answers referencing Reddit, measured 30 September
Every figure is accurate. They measure different quantities. A strategy anchored to any single platform can reprice overnight because a model provider adjusted a weight, and nobody is notified.
Semrush's own analyst disputed the widely repeated explanation, noting that only 34% of Reddit's Google rankings sit in positions 21 to 100, which does not account for a collapse this steep. The cause remains contested. The volatility does not.
Finance Is the Most Volatile Citation Category Tracked
Not an opinion. BrightEdge measures week-on-week citation change across every category, and financial services moves more than any other.
Week-on-week citation change, financial domains
Source: BrightEdge citation tracking. More than half of financial domains see their citation position change week to week, and nine in ten of those changes are losses.
Weekly
Not Annually. Not Quarterly.
An annual AI visibility audit in financial services measures a position that stopped existing in week three. Monitoring cadence is not a service upsell here. It is the minimum viable frequency.
4-8 wks
How Fast Technical Fixes Land
AI crawlers re-index considerably faster than traditional search. Schema deployment and robots.txt corrections can shift citations within two months, which cuts both ways.
How to Write an Answer an Engine Can Lift Without Breaking POJK
Compliance requires disclaimers. Extraction requires clarity. Most institutions solve this by burying the answer under three paragraphs of qualification, and the engine gives up before reaching it.
Disclaimer first, answer buried
Paragraph 1
The information presented here is for general purposes only and does not constitute financial advice. Terms and conditions apply. Individual circumstances vary. Consult a licensed professional before making any decision.
Paragraph 2
Interest calculation methods vary across products and institutions. Regulatory frameworks in Indonesia are governed by the Financial Services Authority under several applicable provisions.
Paragraph 3
A flat monthly rate of 1.2% over 24 months corresponds to approximately 22% effective annually.
Roughly 44% of citations are extracted from the first 30% of a page. The answer is in the last third.
Answer first, compliance intact
Sentence 1, the answer
A flat monthly rate of 1.2% over a 24-month tenor is approximately equivalent to 22% effective annual interest, because flat rates are calculated on the original principal while effective rates account for the declining balance.
Sentence 2, regulatory context
Under OJK transparency requirements, institutions must disclose both figures before an agreement is signed.
Sentence 3, disclaimer
Exact figures depend on repayment structure and individual terms. This explanation is educational and does not constitute financial advice.
Identical compliance posture. The extractable sentence now sits where engines actually read.
Both versions satisfy the same regulatory requirements. Only one of them gets cited.
The Peer-Reviewed Evidence, and What It Says About Regulated Content
Aggarwal and colleagues published the first peer-reviewed study of Generative Engine Optimization at ACM KDD 2024, testing ten thousand queries across nine domains using a benchmark they named GEO-BENCH.
+30-40%
Visibility gain from citing sources, adding quotations, and adding statistics
Zero
Effect from keyword stuffing. Sometimes negative.
Larger
Gains for lower-ranked sites than for already-dominant ones
The finding that matters for finance: the domain closest to regulated financial content, Law and Government, benefited most from adding statistics. Citing regulators and quantifying claims is not merely good compliance practice. It is the highest-leverage GEO tactic available in this category, and the paper measured it.
And the finding that matters for challengers: sites ranking lower in traditional search gained more from GEO than sites already dominating. A regional bank has more to gain here than a national one, which inverts the usual dynamic of search.
What does not work: JavaScript-injected schema. AI crawlers read the server-side HTML head before JavaScript executes, so client-side schema makes your rates, terms and licence numbers invisible.
Six Pillars of Financial GEO
Built for a category where the engines hedge, the regulator watches, and the citations move weekly.
1
Answer-First Architecture Inside Compliance
The disclaimer stays. It moves. Sentence one carries the answer, sentence two the regulatory context, sentence three the required qualification.
- Extractable summaries in the top 30% of every page
- Modular disclosure blocks, pre-approved, reused
- Comparison tables that engines lift intact
- Identical compliance posture, different citation outcome
2
Regulator Citation as a Ranking Signal
The KDD 2024 study found Law and Government content benefited most from statistics and source citation. Financial institutions have both, and rarely deploy either.
- Direct citation of OJK, Bank Indonesia and applicable POJK
- Quantified claims with verifiable figures
- Licence numbers as machine-readable text, never badge images
- Methodology transparency on every calculation
3
Credentialed Authorship and Entity Recognition
Gemini favours brands it recognises as entities. Knowledge Graph presence, verified panels, consistent schema. Most Indonesian financial institutions have none of it.
- Person schema with credentials and licence numbers
- Organization schema with sameAs references
- Knowledge Panel and entity disambiguation work
- Editorial review disclosure on every YMYL page
4
Server-Side Financial Schema
FinancialProduct, LoanOrCredit, InterestRate, MonetaryAmount. Rendered in the HTML head, not injected after page load.
- Nested @graph configuration, not isolated blocks
- Rates updated in structured data without republishing
- FAQPage schema on regulatory and trust queries
- Migration off any JavaScript-injected implementation
5
Weekly Citation Monitoring
More than half of financial domains see citation position change week to week, and 91% of those changes are declines. An annual audit measures a position that no longer exists.
- Per-engine tracking across ChatGPT, Gemini, Perplexity, AI Overviews
- Prompt sets covering information, comparison and validation intent
- Sentiment and reference depth, not only presence
- Alerting on retrieval-weight shifts before they cost a quarter
6
Diversified Citation Portfolio
September 2025 demonstrated what concentration costs. A single platform can lose most of its citation share in six weeks because a configuration file changed.
- Earned media across named Indonesian financial publications
- Regulator-adjacent educational content
- Video and community presence, weighted per engine
- No single source exceeding a defensible share
Three Things About This Market the Global Playbook Misses
Perplexity Punches Seven Times Above Its Global Weight
Roughly 11 to 21 percent of Indonesian AI chatbot share, against a global average nearer 2 to 3 percent. The reason is documented: Telkomsel bundled Perplexity Pro into prepaid and postpaid plans from 28 May 2025.
- Ignoring Perplexity forfeits a fifth of the market
- It weights recency and cross-source corroboration heavily
- Numbered citations make it the most measurable engine
Machine Translation Is Detected and Penalised
Generative engines identify unnatural phrasing and down-rank it during citation selection. Translating an English explainer into Bahasa Indonesia is not localisation.
- Natively written Bahasa, paired via hreflang
- Sahabat-AI, a 70-billion parameter model from Indosat and GoTo, trained on Bahasa, Javanese and Sundanese
- Financial queries mix formal Bahasa, English loanwords and slang in one sentence
The Regulator Is Watching AI Communication
OJK published an AI governance framework for the financial sector in 2025, and POJK 6/2026 places liability for influencer content on the institution rather than the creator.
- Sanctions up to fifteen billion rupiah, borne by the licensed entity
- POJK 22/2023 prohibits superlatives and misleading claims
- What an AI says about your product is becoming your responsibility
Indonesia-specific polling isolating AI use for financial research does not yet exist in a form we can verify. The EY figure of 49% is a global measure across 23 countries, and we present it as a directional proxy rather than a local finding.
Our GEO and AEO Services for Financial Services
Everything a regulated institution needs to be the source these engines cite, in a category where they hedge by default.
AI Visibility Audit for Financial Brands
Baseline citation share across ChatGPT, Gemini, Perplexity and AI Overviews, mapped against the prompts your customers actually use before borrowing, investing or insuring.
Includes the prompts where engines refuse to answer, because those reveal where your competitors cannot reach you either.
Answer-First Content Restructuring
Roughly 44% of citations are extracted from the first 30% of a page. Most financial content buries the answer beneath three paragraphs of disclaimer.
We restructure without weakening a single compliance obligation. The disclaimer stays. It moves.
Financial Schema Engineering
FinancialProduct, LoanOrCredit, InterestRate and MonetaryAmount, rendered server-side in a nested @graph configuration.
Includes migration off JavaScript-injected schema, which AI crawlers cannot read, and which currently makes rates and licence numbers invisible on most Indonesian financial sites.
Entity and Authorship Authority
Gemini favours brands it recognises as distinct entities. Knowledge Graph presence, verified panels, consistent Organization schema with sameAs references.
Paired with Person schema carrying real credentials, because a YMYL page bylined "Editorial Team" carries a structural weakness no link building repairs.
Weekly Citation Monitoring
Financial services is the most volatile citation category tracked. More than half of domains change position week to week, and 91% of those changes are declines.
Annual audits measure a position that expired in week three. We monitor weekly and alert on retrieval-weight shifts before they cost a quarter.
Compliance-Aligned Citation Building
Under POJK 6/2026, sanctions for influencer content fall on the licensed institution. Off-site citation building in this category is a compliance exercise before it is a marketing one.
Earned media across named Indonesian financial publications, regulator-adjacent educational content, and no single source carrying an indefensible share.
to be cited before they decide, use GEO
Why Choose Us for Financial GEO?
Bridging Two Decades of Digital Excellence with Regulated-Sector Discipline
Financial content answers to Google's strictest quality tier, to OJK, and now to engines that hedge on money questions by design. Most agencies have never worked inside one of those constraints. Choosing a GEO agency for regulated finance means choosing one that has read the regulation.
15+
Countries We Operate In
6
AI Platforms Actively Optimized
We Cite the Research, Not Vendor Marketing
The peer-reviewed KDD 2024 study measured what works: source citation, quotations and statistics gain 30 to 40 percent visibility, while keyword stuffing gains nothing. We build to the evidence.
Compliance Is a Design Constraint, Not an Obstacle
The disclaimer stays. It moves. Restructuring for extraction without weakening a single regulatory obligation is the entire craft in this category.
We Monitor Weekly Because Finance Moves Weekly
The most volatile citation category tracked. September 2025 proved a channel can vanish in six weeks because a configuration file changed, with no announcement.
Institutional-Grade Governance
ISO 9001, ISO 14001 and OHSAS 18001 certified. Documentation and change control structured for organisations where every claim needs an audit trail.
Explore Related Services
Financial GEO rests on technical SEO foundations. Neither works alone.
Ready to Be the Source the Engines Trust?
Get a free AI visibility audit showing exactly which financial prompts cite you, which cite an aggregator, and which the engines refuse to answer at all. Contact our team to get started.
Request Your Free AI Visibility Audit
Frequently Asked Questions About GEO and AEO for Financial Services
Why do AI engines refuse to answer some financial questions?
Because there is a line between information and advice, and the engines enforce it. Ask how effective interest is calculated and you get a cited answer. Ask whether you should take a specific loan and you get a refusal, a disclaimer, and a suggestion to consult a licensed advisor. Financial institutions can only occupy the first territory, and the entire GEO strategy for this sector follows from understanding that boundary.
If consumers do not trust AI for financial advice, why does GEO matter?
Because they consult it anyway. EY found 49% of consumers used AI for savings and investment decisions in the past six months, across 18,152 respondents in 23 countries. YouGov found only 20% express any trust in AI within financial services, the lowest of any sector. TD Bank found 78% use AI tools while just 18% would trust it with a recommendation. People ask the machine, then look for a credible source to confirm what it said. Being that source is the opportunity.
How do we get cited when compliance requires disclaimers everywhere?
By moving the disclaimer rather than removing it. Roughly 44% of citations are extracted from the first 30% of a page, so an answer buried beneath three paragraphs of qualification never gets read. Sentence one carries the answer. Sentence two carries the regulatory context. Sentence three carries the required disclaimer. Identical compliance posture, entirely different citation outcome.
What does the peer-reviewed research actually say about GEO?
Aggarwal and colleagues published the first peer-reviewed GEO study at ACM KDD 2024, testing ten thousand queries across nine domains. Citing sources, adding quotations and adding statistics improved visibility by 30 to 40 percent. Keyword stuffing produced zero effect and sometimes a negative one. Crucially, lower-ranked sites gained more than already-dominant ones, and the Law and Government domain, the closest proxy for regulated financial content, benefited most from statistics.
Did Reddit's citation share really collapse in September 2025?
Yes, and how far depends entirely on who measured it. Spotlight tracked a fall from 14.29% to 0.21% of all cited sources. SEMrush tracked roughly 60% to 10% of prompt responses. An agency study reported to RBC Capital said 29.2% to 5.3%. PromptWatch measured 9.7% to 2% of ChatGPT answers. All four are correct. They measure different quantities. The cause remains contested, with Semrush's own analyst disputing the widely repeated explanation. What is not contested is that a channel can lose most of its value in six weeks because a configuration changed, with no announcement.
How often should we monitor AI citations in financial services?
Weekly. BrightEdge tracks financial services as the most volatile citation category, with more than half of financial domains seeing their position change week to week, and 91% of those changes being declines. An annual AI visibility audit measures a position that stopped existing in week three.
Does JavaScript-injected schema work for AI crawlers?
No. AI crawlers read the server-side HTML head before JavaScript executes. Schema injected after page load renders correctly in a browser and validates in testing tools, but the crawler finds nothing. On a financial site this means rates, terms and licence numbers are legally displayed to a human and completely invisible to every engine deciding what to cite.
Why does Perplexity matter more in Indonesia than globally?
Because Telkomsel bundled Perplexity Pro into prepaid and postpaid data plans from 28 May 2025, the first connectivity and AI bundle in Indonesia. Perplexity now holds roughly 11 to 21 percent of Indonesian AI chatbot share against a global average nearer 2 to 3 percent. A GEO strategy copied from a Western agency underweights it badly, and Perplexity weights recency and cross-source corroboration differently from ChatGPT.
Are we liable if an AI gives someone bad advice about our product?
The regulatory direction suggests increasing institutional responsibility. POJK 6/2026 places sanctions for financial influencer content on the licensed institution rather than the creator, with penalties reaching fifteen billion rupiah. OJK published an AI governance framework for the financial sector in 2025. What an AI says about your product is becoming your concern, which makes citation accuracy a compliance matter and not only a marketing one.
How long until GEO work shows results?
Technical changes such as server-side schema deployment and robots.txt corrections can shift citations within four to eight weeks, since AI crawlers re-index considerably faster than traditional search. Building genuine third-party consensus takes longer, typically three to six months of consistent presence. In financial services, both timelines sit inside a category where positions change weekly, which makes monitoring as important as the work itself.
Does GEO replace our SEO investment?
No. AI engines rely on search crawlers to reach your content before anything else can happen. A financial site Google cannot index properly will not be cited well by ChatGPT either, however good the structured data.
Fintech SEO is the foundation. GEO sits on top of it and collapses without it.
Should we translate our content into Bahasa Indonesia?
Translation is the wrong approach. Generative engines detect unnatural phrasing and down-rank machine-translated content during citation selection. Natively written Bahasa Indonesia paired with English through hreflang is required, reflecting how Indonesians actually search, mixing formal Bahasa, English loanwords and casual abbreviation inside a single financial query. This also positions the institution for Sahabat-AI, the 70-billion parameter model from Indosat and GoTo trained on Bahasa Indonesia, Javanese and Sundanese.
