How to rank on ChatGPT: 7 tactics that work

Can you rank on ChatGPT?

You cannot rank on ChatGPT the way you rank on Google. There is no position one, no results page, no click-through rate to chase. ChatGPT does not keep a sorted list of websites. It generates each answer by synthesising its training data and, when Browse is active, live web crawls through Bing's index. So the literal question is the wrong one.

The question that pays is different: when someone asks ChatGPT for the best provider in your category, does it name you? That is the functional equivalent of ranking, and it is winnable. When the model recommends a specific business in a commercial query, that business has ranked in the only sense that matters. They got the customer. The tactics below are how you become the business ChatGPT names, in order of how much they move the needle.

1. Publish authoritative content on Medium

Medium is ChatGPT's number one citation source at 14.3% of citations. The model was trained heavily on Medium and continues to index it when Browse is active, which makes it the highest-impact single move you can make. It works because it gives ChatGPT exactly what it likes to quote: clear, authored explanations that look current, structured, and attributable to a real person or company. A weak Medium article does little. A strong one does several jobs at once. It defines the topic in language the model can reuse, reinforces your brand as an entity through the byline and author bio, and creates a fresh source that lines up with what is already on your site. Write with recommendation intent. Answer one buyer question cleanly, make your point of view obvious, and pack in enough specific fact that the model can quote you without inventing the missing detail.

The mechanics matter here. Articles should be at least 800 words, with 1,200 to 2,000 the range where citation frequency peaks. Use clear H2 and H3 structure, put your business name in the author bio and naturally in the body, and publish under a company or founder byline, never anonymously. Match each article to a real question buyers ask in your category, because generic thought leadership performs worse than content built around a specific query. Aim for a monthly cadence at minimum. A single strong article on your core topic builds a citation signal in 4 to 8 weeks, and a business with 12 articles across distinct questions builds far more authority than one with three, even at equal quality. Consistency compounds; raw volume does not.

2. Implement FAQPage schema

FAQPage JSON-LD tells AI crawlers exactly which questions your content answers and what the answers are. When your FAQ holds a direct, factual answer to a question buyers ask, ChatGPT's Browse feature can pull that answer straight into a response. Add a JSON-LD block in your page head with `@type: "FAQPage"`. Each question goes in the `name` field of a `Question` object, and the answer goes in `acceptedAnswer.text`. Keep answers under 60 words for the highest extractability, aim for 4 to 6 pairs per page, and run Google's Rich Results Test to confirm the block is valid before you publish. Deploy it on your homepage, every service page, and any content page that answers buyer questions directly.

3. Build consistent brand entity recognition

ChatGPT uses entity recognition to identify businesses. If your name, location, description, and service area read the same across multiple trusted sources, the model can treat you as one coherent entity and recommend you with confidence. The most explicit lever is Organization schema in JSON-LD. Include your canonical business name, your URL, a one to two sentence description, and `sameAs` links to LinkedIn, Twitter/X, and any other verified profiles. The description in your Organization schema should match your LinkedIn company description and your Google Business Profile description word for sense. Inconsistency across sources is the most common reason a model that clearly understands your category still cannot place you in it. When the facts disagree, the model hedges, and a hedged model recommends the competitor whose facts agree.

4. Get cited in third-party content

A mention of your business in someone else's article, review, or comparison is among the strongest signals ChatGPT uses. It is independent proof that you exist and matter, which is harder to fake than anything on your own domain. Pursue guest articles in industry publications where the byline or copy names you, press coverage in trade or local business media, comparison content that lists the top providers in your category, and podcast appearances whose transcripts include your business name. Each independent source that names you builds the entity authority the model leans on to construct a confident recommendation. Third-party mentions take 4 to 12 weeks to surface depending on how fast each source gets indexed, so start the outreach early and let it accumulate.

5. Publish LinkedIn articles consistently

LinkedIn articles, the long-form format rather than ordinary posts, are indexed by ChatGPT and weighted as professional authority signals. They appear in 5 to 7% of B2B ChatGPT citations, less than Medium but doing a different job: entity verification. When the model sees a steady stream of LinkedIn writing from a company and its founder across one topic area, it builds confidence that the business is real, credible, and relevant. Run a consistent series on your core topic, one article every two weeks, and you build the kind of topical expertise ChatGPT draws on for professional-context queries. Pair it with Medium rather than treating it as a substitute, because the two signals reinforce each other.

6. Participate in industry forums and Q&A platforms

ChatGPT draws on Quora, Stack Exchange, and other Q&A platforms for specific factual questions. Answer questions there with accurate, useful detail, and mention your business only where it is genuinely relevant. The same quality bar applies as everywhere else: a specific, helpful answer builds a citation signal, while a promotional one is ignored or penalised. Treat these platforms as places to be useful in public, not places to advertise.

7. Monitor what ChatGPT says about your business

Monitoring matters because ChatGPT tells you what is missing long before your analytics do. If the model can describe your category but cannot name you, the problem is usually weak entity coverage or thin third-party reinforcement. If it names you but frames you vaguely, you have a positioning or proof gap. If it cites a competitor in sharper language than it uses for you, that is a direct blueprint for your next piece of content. None of this shows up if you only track pageviews. Run a fixed set of queries monthly, save the responses, and compare the wording, not just whether you were mentioned. The exact phrasing tells you whether the engine sees you as generic, specific, current, local, premium, or uncertain, and those adjectives shape every recommendation that follows.

Run these queries monthly: "What is the best [your service] in [location]?", "Who are the top [your category] providers?", and "[Your business name], what do you know about them?" Record the answers. The gaps point to the specific content interventions you need. Share of voice across 30 target queries, tracked month over month, is the clearest single measure of whether your programme is working. If you do not know what ChatGPT currently says about your business, you are building blind.

How ChatGPT selects its sources

ChatGPT pulls from two places: its training data, which is heavily weighted toward high-quality public web content, and, when Browse is active, real-time crawls through Bing's index. For any live query, Bing is the gating factor, which means your site needs to be indexed in Bing and not only Google. Check your presence in Bing Webmaster Tools and submit your sitemap if it is missing. Across both training-data and Browse responses, the platforms that appear most often in citations are Medium, LinkedIn, Wikipedia, GitHub, and academic publishing. Of these, Medium at 14.3% and LinkedIn are the most actionable for most businesses. A Wikipedia presence is possible if you meet the notability bar with neutral, verifiable sourcing, but pages that read like marketing get deleted. This is the mechanism that turns the whole field into a discipline, which is the subject of our answer engine optimization guide and worth reading alongside this piece.

How long it takes to get cited

Set expectations against the actual timeline. Schema and technical fixes are visible to crawlers within days. Medium content takes 4 to 8 weeks to build a citation signal. Third-party mentions take 4 to 12 weeks depending on indexing speed. A consistent improvement in citation share takes 3 to 6 months of steady execution. In practice, weeks one to four are foundation work: schema goes live and the first Medium articles publish, with little visible change. Months two and three bring the second and third articles, and the first citations may appear for low-competition queries. Months four to six show citation frequency climbing as cumulative authority builds and higher-competition queries start to surface you. From months six to twelve the compounding becomes measurable and you appear consistently in answers to your core customer queries. The businesses that start now sit 6 to 12 months ahead of the ones that wait for ChatGPT citation to become an obvious priority.

How twohundred would approach this

If we were scoping this for an operator, we would not start by writing 12 Medium articles. We would start by running the 30-query monitoring set so we knew exactly what the model already says, then fix entity consistency first, because that is the cheapest change with the biggest effect on whether you get named at all. Schema is a one-time development cost. Medium publishing costs nothing in-house, or roughly £200 to £400 per article with a writer. Only then would we build the content cadence, sequenced against the specific queries where you are losing to a competitor. The order is the work. Most businesses spend on content before they know which content the model is missing, and the spend compounds slowly because of it. If you want a second pair of eyes on your current stack or a scoped first build, twohundred can walk through what you have and what is worth building first.

Frequently asked questions

How much does it cost to get cited by ChatGPT?

The core programme of FAQPage schema, Medium publishing, and LinkedIn articles needs either time or a modest content budget. Schema is a one-time development cost. Medium publishing costs nothing done in-house, or around £200 to £400 per article if you hire a writer. The monitoring itself is free. The real cost is consistency over three to six months, not any single line item.

Does having a Wikipedia page help?

Yes, significantly. Wikipedia is among ChatGPT's most trusted sources for entity information, so a neutral, verifiable page citing reliable sources gives much stronger entity recognition than no page at all. The catch is that Wikipedia has strict notability requirements, and articles that read like marketing are deleted. It only helps businesses with genuine, independently documented notability.

Will being cited on Perplexity or Google help with ChatGPT too?

Partially. Some signals, particularly domain authority and third-party citations, transfer across engines. But the platforms differ. Medium is the dominant ChatGPT signal, while Reddit is the dominant Perplexity signal. Building for every engine at once requires a deliberate multi-platform approach rather than assuming one effort covers all of them.

What content topics work best for ChatGPT citation?

Specific buyer questions about your category, answered directly and factually. "What should I look for when choosing an accounting firm?" or "How much does dental implant treatment cost?" outperform generic thought leadership every time. The more precisely your content matches a question buyers actually ask, the stronger the citation signal it builds.