How do we show up in AI-generated answers when buyers use ChatGPT or Perplexity instead of Google?
Search behaviour is changing. Buyers who used to type a query into Google and click through to your website now ask ChatGPT or Perplexity the same question and read the answer in the interface. If you are not cited in that answer, you are invisible — regardless of your SEO ranking. This is a new problem that requires a different kind of thinking.
How AI search changes the B2B marketing equation
When a buyer types a query into Google, your SEO ranking determines whether they see you. When a buyer asks the same question to ChatGPT, Perplexity, or Gemini, an AI model decides what sources are relevant and synthesises an answer. Your website might rank first in Google and be completely absent from the AI answer. This is a new visibility problem — and it operates on different rules than traditional SEO.
AI models like ChatGPT and Perplexity retrieve information from two sources: their training data (what they learned before their knowledge cutoff) and real-time web search (what they retrieve when answering a specific query). For B2B queries, real-time retrieval is particularly important because market conditions, vendor landscapes, and product features change frequently. Your content must be publicly indexed and structured for retrieval.
SEO optimises for a ranking algorithm that prioritises relevance, authority, and experience signals. AI retrieval optimises for factual accuracy, source credibility, and content structure. A page that ranks first in Google because it has good backlinks and high engagement metrics is not guaranteed to be retrieved by an AI model answering a specific question. The signals overlap, but they are not the same.
AI models give significant weight to content that is cited by other authoritative sources. A company that is mentioned in a Gartner report, cited in a TechCrunch article, and referenced on G2 is more likely to appear in an AI-generated answer than a company with an excellent website that nobody else cites. This means AEO is partly about your own content and substantially about your presence in third-party sources.
Most AEO content and strategy is currently written for US and European markets. Indian B2B companies have a genuine opportunity to be first-movers in building AI search presence for India-specific B2B queries — 'best IT services company for healthcare in India,' 'B2B SaaS compliance with DPDP Act,' 'manufacturing supplier for aerospace in Bengaluru.' These queries are being asked and the answers are currently thin. Being the clear, structured, authoritative answer to an India-specific query is achievable with modest investment.
The question map: L1 vs L2
L1 questions describe the symptom. L2 questions locate the real cause — and point toward decisions that actually fix it.
Understand what AI models actually retrieve — and what they don't
AI models retrieve content that is: publicly indexed, clearly structured, factually specific, and cited by other authoritative sources. They retrieve less effectively: gated content, overly promotional language, content with no clear factual claims, and content that has no external citation. The first step is auditing your existing content against these criteria.
Enterprise SaaS buyers frequently use AI search for queries like 'best project management software for engineering teams' or 'comparison of Salesforce vs HubSpot for mid-market B2B.' These are the queries where you must appear. The AI model answering these queries pulls from review platforms (G2, TrustRadius), analyst reports, and high-authority content. Your strategy: maximise your presence on review platforms, ensure your category positioning is clearly stated on your website and in public documentation, and build content that specifically addresses comparison queries in your category.
AI models like Perplexity and ChatGPT heavily weight analyst sources — Gartner, Forrester, HfS, Everest Group — when answering queries about IT services vendors. If you are mentioned positively in an analyst report, there is a meaningful probability that mention will appear in an AI-generated answer to a relevant query. This is the most direct argument for analyst relations investment: it is no longer just about influencing human buyers — analyst coverage now directly influences AI-generated shortlists.
AI models retrieve structured, factual data more reliably than narrative marketing content. For manufacturing companies, this means your product datasheets, specifications, compliance certifications, and application notes need to be publicly accessible, well-structured, and clearly attributed to your company. A datasheet that lists material properties in a table format is more likely to be retrieved accurately by an AI model than a product page that describes the same properties in marketing prose. Review your technical documentation library with this in mind.
Pharma procurement teams increasingly use AI search to get quick answers to regulatory compliance questions — 'what does GDP compliance require for a cold chain logistics provider?' or 'which CROs are qualified for Phase II oncology studies in India?' If you publish clear, accurate regulatory guidance content that is specifically relevant to your buyers' questions, there is a meaningful probability of being cited. The content needs to be technically accurate, publicly indexed, and clearly structured — not marketing language, not gated.
Build citable content — content with specific, verifiable claims
The content that AI models cite most often is content with specific, verifiable claims: data points, case study outcomes, named certifications, defined methodologies. Vague claims — 'industry-leading,' 'best-in-class,' 'comprehensive solution' — are not citable. A case study that says 'reduced implementation time by 40% for a 500-person manufacturing company in Pune' is citable. A case study that says 'helped our client achieve significant improvement' is not.
Earn third-party citations — the most important AEO signal
Being mentioned in sources that AI models consider authoritative increases your citation probability more than any on-site content change. The most valuable third-party citation sources for B2B companies are analyst reports, industry publications, review platforms, and high-authority news coverage. Each mention in these sources is a citation that increases your likelihood of appearing in AI-generated answers.
Optimise your structured data and entity presence
AI models build a model of your company as an entity — not just a website. They draw on your Wikipedia presence (if any), your LinkedIn company page, your Crunchbase or Tracxn profile, your Google Business profile, and structured data on your website. Ensuring these sources are consistent, complete, and accurate improves the reliability with which AI models describe your company.
Measure your AI search presence — and track it over time
AI search presence is measurable, but the measurement tools are less mature than traditional SEO tools. The practical approach is a combination of direct testing and structured monitoring. Direct testing means regularly querying AI models with the questions your buyers would ask and checking whether you appear. Structured monitoring means tracking mentions in the third-party sources that AI models cite.
Real-world examples
How B2B companies across India and globally have navigated this decision.
HubSpot's content strategy — building the most comprehensive public library of marketing, sales, and CRM education in the world — was originally designed for SEO. It works equally well for AI retrieval because it is structured, factual, and extensively cited. When an AI model answers a query about CRM software for mid-market companies, HubSpot appears because: it is mentioned in thousands of reviews with specific use cases on G2 and TrustRadius, it is cited in analyst reports, it is referenced in hundreds of comparison articles across high-authority domains, and its own content is highly structured and factually specific. The lesson is not that you need HubSpot's content volume — it is that their content strategy was always about being citable, not just rankable.
A Bengaluru-based compliance SaaS company identified that their target buyers — CISOs and legal teams at SaaS companies — were increasingly using AI search to understand their obligations under India's Digital Personal Data Protection Act. They published a detailed, structured, factually accurate guide to DPDP Act compliance requirements — not marketing content, but a genuine compliance reference document. Within six months, queries like 'DPDP Act requirements for SaaS companies in India' on Perplexity were returning their guide as a primary source. Their organic traffic from this single piece exceeded their previous best-performing SEO content because the structured, factual format was highly retrievable by AI models in addition to ranking in traditional search.
A mid-sized Indian IT services company conducted a test: they queried Perplexity with ten questions their target buyers would ask about IT services vendors — questions about cloud migration, AI implementation, and digital transformation. They appeared in answers for two queries; their three main competitors appeared in answers for seven to nine queries each. The difference was a citation chain: competitors had Gartner and Everest Group mentions, G2 reviews, and regular coverage in CIO-focused publications. The company built an 18-month programme — analyst relations, a structured G2 review campaign, and a PR programme targeting CIO publications. At the 18-month mark, they reran the same ten queries. They appeared in seven of ten answers. No change had been made to their website content; all the improvement came from third-party citation building.
When the logic works — and when it breaks
- Key product and company information is publicly accessible, clearly structured, and factually specific
- Analyst relations, review platforms, and industry publication coverage are treated as AEO investments
- Content contains specific, verifiable claims — not promotional language
- Entity data is consistent and complete across LinkedIn, Crunchbase, website, and directory listings
- AI search presence is tested monthly with buyer-relevant queries across multiple AI models
- India-specific queries are treated as a differentiated opportunity, not an afterthought
- Key product information is behind login walls or in PDFs that AI models cannot retrieve
- Website content is written in promotional language with no specific, citable claims
- Third-party citation building — analyst relations, review platforms, PR — receives no marketing investment
- AEO is treated as an SEO task rather than a visibility strategy that requires third-party presence
- No regular testing of how AI models describe the company and what queries they surface for
- Entity data is inconsistent across platforms — different product names, category descriptions, or founding information
Your move
Open Perplexity and ChatGPT right now. Type the query your best customer would have asked before they found you — something like 'best [your category] for [your target industry] companies in India.' Look at the answer carefully: are you mentioned? Are your competitors? What sources does the AI cite? That gap between your position and your competitor's position is your AEO starting point.
Then check your G2 or TrustRadius page. If you have fewer than 15 reviews, or if your reviews are generic rather than specific, that is your first investment. AI models cite review platforms heavily for software category queries. Ten specific, detailed reviews from real customers with named outcomes will improve your AI search presence more quickly than any on-site content change.
Every month, one hard B2B marketing problem.
First principles thinking. Real India context.