The Indian Directory AI Citation Study: How Platform Search Ranking Drives LLM Citations

Research Purpose

A fundamental question for Indian AEO strategy: "Is it more important to rank on Practo/JustDial, or to optimise our own website for AI?"

This study addresses this empirically, measuring the correlation between directory ranking and LLM (ChatGPT, Perplexity) citation rates for Indian local businesses across 4 categories.

The hypothesis: Indian directory ranking is the primary driver of ChatGPT and Perplexity citations for most Indian local business queries, because these AI systems browse search results that are dominated by Indian directory pages.

Status: this is a research protocol. The hypothesis, the "expected" pathway, and the implications sections below describe what the study is designed to test — none of it is a confirmed finding yet.

Why this question matters to a real decision

This is one of the highest-stakes prioritisation questions in Indian AEO, because the two answers point to genuinely different investment sequences. If directory ranking is the dominant driver of ChatGPT and Perplexity citations, a healthcare business should focus its limited AEO budget on climbing Practo's category rankings before spending heavily on website schema and content — because the AI systems are mostly reading Practo, not the clinic's own site. If website quality matters just as much, that sequencing is wrong, and a business skipping its own website in favour of directory optimisation is leaving half the opportunity on the table.

Most current Indian AEO advice — including some of Angryturtle's own general guidance — leans toward "invest in your website's schema and content," largely because that's the transferable wisdom from US AEO practice, where Yelp and similar directories don't dominate AI citation the way Practo, Zomato, and JustDial plausibly do in India. This study exists specifically to check whether that imported assumption actually holds for the Indian market, or whether it's sequencing Indian businesses' AEO investment incorrectly.

Study Design

Categories studied: Healthcare (Practo as primary directory), Restaurants (Zomato), Real estate (99acres), General services (JustDial)

Sample per category: 50 businesses per category — stratified across directory ranking positions (1–5, 6–10, 11–20, 20+)

Measurement: Monthly, for each business:

1. Current directory ranking for 5 standardised category + city queries on that category's primary directory
2. ChatGPT citation rate for the same 5 queries
3. Perplexity citation rate for the same 5 queries
4. Business website ranking for the same 5 queries on Google

Why this hasn't been measured before

Answering this question requires tracking two independent ranking systems for the same set of businesses at the same time — a business's position within a directory's own internal search (Practo's ranking of dermatologists in Koramangala, say) and that same business's citation rate on ChatGPT and Perplexity for the equivalent query. Neither of those numbers is available through an API; directory internal rankings have to be observed by running the same query on the directory itself, monthly, and cross-referenced against separately observed AI citation results for the same businesses.

The harder part isn't the mechanics of running the queries, it's building genuinely comparable query sets across four structurally different categories — a Practo search for a dermatologist doesn't behave like a Zomato search for a restaurant, and a study that used different query logic for each category would make the cross-category comparison unreliable. Designing four category-specific but methodologically equivalent query protocols, then running all of them monthly against Google, ChatGPT, and Perplexity for 200 businesses, is a sustained data-collection commitment nobody appears to have made publicly for the Indian market before this.

The Citation Pathway Model

The pathway being tested:

User query → ChatGPT/Perplexity browses Bing/Google → Bing/Google returns directory pages → AI crawls directory pages → AI extracts business data from directory listing → AI cites business in response

If this pathway is primary: Directory ranking is the critical variable for ChatGPT and Perplexity citations. Website optimisation is secondary.

Alternative pathway:

User query → ChatGPT/Perplexity browses Bing/Google → returns business websites and directory pages → AI crawls both → AI cites the more informative source

If this pathway is equally primary: Website quality and ranking are as important as directory ranking for LLM citations.

Measurement Methodology

Directory ranking measurement: For each business, a standardised set of queries is run on the primary directory (e.g., on Practo for healthcare: "dermatologist in Koramangala"). The position of the business listing in Practo's category search results is recorded.

ChatGPT citation measurement: The same queries are run in ChatGPT Plus (with browsing, in fresh conversation sessions). For each query, analysts record: was the business cited? Was the citation source the directory page or the business's own website?

Perplexity citation measurement: Same queries in Perplexity. The source chips in Perplexity's response make source attribution more directly observable than in ChatGPT — Perplexity explicitly shows which URL it cited.

Source attribution analysis: For citations that do occur, the source URL is recorded. This allows calculation of what proportion of ChatGPT citations for healthcare businesses come from Practo pages versus business websites.

What would invalidate this study

Causal direction: Correlation between directory rank and citation rate doesn't establish causality. Businesses with high directory rank may also have better Google reviews, better website content, and more directory presence generally — all of which could independently affect citation rates. A business that dominates Practo's category rankings is also likely investing in its overall online presence, and disentangling "the directory rank itself caused the citation" from "this is just a well-run business that ranks well everywhere" is genuinely hard with an observational design.

Platform variability: ChatGPT and Perplexity may use different search engines for different queries, and may cache results. Monthly measurement and multiple queries per business mitigate but don't eliminate this.

Platform evolution: Perplexity and ChatGPT browse mechanisms may change during the study period. The study's source attribution analysis would reflect such changes in real time, which is useful, but it also means a finding from month 3 of the study might not hold by month 10 if either platform meaningfully changes how it retrieves sources.

The moderating-variable controls in the analysis — directory review count, Google review count, and website search ranking for the same queries — address the most obvious confounds directly, but as with any observational study, they reduce rather than eliminate the risk that the correlation reflects general business quality rather than directory ranking specifically.

How a business can check this for itself

You can get a rough directional read on your own citation pathway without running a four-category study. Take your five most important category + city queries and check three things for each: where you currently rank inside your category's primary directory's own search (Practo, Zomato, 99acres, or JustDial, depending on your category), where your own website ranks on Google for the same query, and whether ChatGPT or Perplexity cites you at all when you ask the equivalent natural-language question — and if so, which URL it cited.

If your directory ranking is strong but your website ranking is weak, and AI citations still name you, that's a real, specific data point suggesting the directory pathway is doing the work for your business. If the opposite pattern holds — weak directory ranking, strong website, and you're still getting cited from your own site — that tells you your website is carrying more weight than the directory-dominant hypothesis would predict. Either way, checking Perplexity specifically is worth doing first, since it shows its source citations directly rather than requiring you to infer the source the way you sometimes have to with ChatGPT.

Correlation Analysis

Primary correlation (hypothesis test): Spearman rank correlation between directory ranking position (1 = top ranking) and ChatGPT citation rate (% of queries generating citation), and separately between directory ranking position and Perplexity citation rate.

Expected direction: negative correlation (lower rank position number = higher citation rate). Expected magnitude: strong (r > -0.6) if directory ranking is primary.

Secondary analysis — source attribution: For citations that do occur: what proportion cite the directory page vs the business's own website?

If directory page citations dominate (>60%): confirms directory ranking as primary citation driver. If website citations dominate or split evenly: suggests direct website optimisation is equally important.

What US-market research doesn't tell us here

US AEO guidance generally treats a business's own website as the primary citation target and directories as secondary, because the US directory landscape — Yelp, TripAdvisor, a scattering of category-specific sites — doesn't dominate search results for local queries the way Practo, Zomato, 99acres, and JustDial plausibly do in India. Applying US-derived AEO sequencing advice to an Indian business without checking whether the underlying pathway assumption holds is exactly the risk this study is designed to catch; no specific US citation-pathway study is referenced here as evidence for the Indian market, because none has been verified as transferable.

Implications of Expected Findings

If directory ranking is the primary driver (expected): AEO investment sequencing for Indian businesses would look like directory presence and ranking first (Practo for healthcare, Zomato for food, JustDial universal), then website schema and content as a secondary amplifier. Current typical Indian AEO advice that prioritises website schema over directory presence may be sequencing the investment incorrectly.

If website ranking is equally important: This would confirm that Indian businesses need to invest simultaneously in directory presence and website SEO/AEO — neither alone sufficient for ChatGPT and Perplexity citation coverage.

The practical implication either way: Both directory presence and website quality plausibly matter to some degree. The study is designed to quantify the relative weight, which is what actually matters for prioritising a limited marketing budget.

Data Transparency and Publication Protocol

Pre-registration: Study hypotheses and methodology are documented before data collection to prevent post-hoc hypothesis revision.

Public output: Correlation coefficients, source attribution percentages, and investment sequence implications will be published once the monthly measurement series across all four categories completes. Individual business data is not published.

Replication: The methodology is described in sufficient detail to be independently replicated. Angryturtle encourages academic collaboration to independently validate findings.

FAQ Section

Q: How can my business participate in this study? A: Angryturtle's client businesses may be included in the research sample (with their consent). Non-client businesses can participate through Angryturtle's research collaboration programme.

Q: Will this study change Angryturtle's AEO recommendations? A: Yes — if the study finds that directory ranking is a stronger driver than website ranking for ChatGPT and Perplexity citations, Angryturtle will update its recommended investment sequence for Indian businesses to prioritise directory optimisation ahead of website AEO.

Q: Should I neglect my website until this study publishes? A: No. Even under the directory-dominant hypothesis, the study's design still expects website citations to occur in some proportion of cases — the question is relative weight, not exclusivity. Both remain worth investing in while the data is collected.

Follow the Indian Directory AI Citation Study →

Internal links: Top 7 Indian Directories listicle · AEO Services · Generative Engine Optimization · Citation Aggregators India glossary · Citation glossary · Citation Frequency glossary · AI Citations: ChatGPT, Perplexity, Gemini blog · NAP Inconsistency & AI Entity Confidence study · Perplexity Optimization India · Get Cited by ChatGPT · Product: Citations & NAP · Industries: Real Estate AI Search · Compare: AEO vs SEO