How to Get Your Firm Recommended in Google AI Mode, AI Overviews and ChatGPT

To get your firm cited in AI search, make three things true: AI can clearly tell who you are and what you're best at, your website answers buyer questions in short, quotable paragraphs, and other trusted sites say the same good things about you that you say about yourself. Do those consistently, with a clear entity, answer-first pages, third-party proof and original data, and AI starts naming you. Skip any one of them, and it may name a rival instead.

Here's the scene this is about. A startup founder in Mumbai opens ChatGPT and types: “best tax advisor for startups in Mumbai.” Three firms get named. Yours isn't one of them, but the firm across town, the one you beat on expertise every single time, is. No ad ran. No one clicked ten blue links. The shortlist was written before your firm ever got a chance to pitch.

That moment is now the first round of buying for professional services. This guide explains how to get into that shortlist, what AI actually looks for, what to fix first and what to stop wasting money on. It's written for law firms, CA and consultancy practices, and other expertise-led firms, because that's where the stakes are highest. One named recommendation can be worth a client relationship that lasts a decade.

Why Has AI Search Changed How Clients Choose a Firm?

AI search has changed buying because the answer now arrives before the shortlist. Instead of scanning ten websites, a buyer reads one AI summary that names two or three firms and usually contacts one of them. If your firm isn't in that answer, you're not being rejected. You're not being considered at all.

The uncomfortable part? The buyer never knows you existed. There's no page-two ranking to climb from, no “we almost saw them.” AI answers compress the whole market into a handful of names, and the firms named tend to be the ones AI can describe confidently. They aren't necessarily the best firms, just the clearest ones.

This is also why AI visibility sits on top of solid SEO foundations, rather than replacing them. AI systems still pull heavily from pages Google and Bing already trust. If your site is weak, your rankings are poor or your content is thin, AI has little solid material to work with, no matter how clever the prompt tricks sound.

What's the Difference Between AI Mode, AI Overviews and ChatGPT?

They're three different doors into the same shortlist. AI Overviews are the AI summary sitting on top of a normal Google search. AI Mode is Google's full conversational mode, more like a chat than a traditional results page. ChatGPT answers using its training, live browsing and third-party sources. The sources differ slightly, but what earns a mention is broadly similar across all three.

That matters because firms waste months optimising for ChatGPT as if it's a separate universe. It isn't. ChatGPT can draw on Bing's index, directories and review platforms. Google's AI draws on Google's index and knowledge graph. Both need the same raw materials: a clear business identity, quotable answers and outside confirmation. Build those properly, and you improve your chances of appearing across all three.

How Does AI Decide Which Firms to Name?

AI names the firms it can describe without guessing. It brings together information from your website, Google Business Profile, directories, reviews and articles about your firm to build a picture of your business. When that picture is consistent and specific about who you serve, what you do, where you operate and why clients should trust you, AI has stronger grounds to recommend you. When the information is vague or contradictory, it may choose a clearer alternative.

Think of it as a very fast, very literal researcher. It doesn't simply recognise that your firm is excellent. It looks for information it can safely repeat. “A 14-partner firm advising startups on GST and fundraising compliance in Mumbai” gives it something concrete. “A leading provider of holistic solutions” could describe almost anyone, so it tells the reader very little.

Consistency is where many firms fall down without realising it. Your website lists one founding year, while a directory lists another. Your LinkedIn profile includes services you stopped offering in 2023. A partner's surname is spelt differently across profiles. People may overlook these discrepancies, but AI systems can interpret them as uncertainty. Uncertainty makes a recommendation harder to justify.

This gap between how good a firm actually is and how clearly the internet can prove it is what Lucien calls the authority gap. Closing it requires consistent information across your online presence. It's unglamorous work, but it can have a significant effect on how confidently your firm is represented.

What Is Query Fan-Out, and Why Must One Page Answer Ten Questions?

Query fan-out is what happens when an AI system breaks a buyer's question into several smaller questions. A single query may lead to searches about cost, specialisation, location, reviews and how to choose a provider. A page that answers several of these related questions gives the system more useful material to work with. A page that answers only the headline question may contribute just one small part of the final response.

Take that Mumbai prompt again. Behind “best tax advisor for startups in Mumbai,” the system may also be looking for answers to questions such as: What do they charge? Do they handle fundraising, or only compliance? What do clients say? Who leads the practice?

Your rival's website may answer most of these questions on clear, well-organised pages. Yours might have a polished homepage and a services page that hasn't been updated in years. The AI didn't necessarily reject your firm. It simply had less information from which to build a useful answer.

This is why structure matters. Use question-based headings, give a direct answer in the first 40 to 60 words under each heading, then add supporting evidence, examples and nuance. This isn't about writing for robots. It's about answering a client's question the way a good partner would: conclusion first, reasoning afterwards.

How Should Your Firm's Pages Be Structured So AI Can Quote Them?

Structure every important page around a direct answer, clear question-based headings, short paragraphs, useful comparison tables, named experts and appropriate structured data. The goal is to make the information easy for both people and machines to understand. Pages that communicate their main points clearly are easier for AI systems to interpret and potentially cite.

The Five Structural Fixes That Matter Most

Answer first, always. The opening paragraph of the page, and of every major section, should state the main conclusion. Use the rest of the section to explain and support it.

One idea per heading. If a section answers two separate questions, consider splitting it. Clear headings help readers find the information they need and help AI systems interpret individual sections.

Add the numbers. Include relevant fee ranges, timelines, team sizes, years of experience and matter types where you can substantiate them. Specific details are more useful than broad adjectives.

Mark it up. Use appropriate Organisation, Service, FAQPage and Person schema where relevant. The structured data must accurately reflect the visible content. Never use schema to claim something the page itself doesn't support.

Name your experts. Use genuine author bylines, professional credentials and short biographies. Anonymous content gives readers less information with which to assess the expertise behind it.

This is the core of answer engine optimisation services done properly. It isn't simply a plugin or a file upload. It involves improving how your most important pages answer the questions potential clients ask. The work takes time, but a competitor cannot easily replicate a well-developed body of useful, specific content overnight.

Why Do E-E-A-T, Named Authors and Original Data Decide Who Gets Cited?

AI systems need sources they can use with confidence. Google's E-E-A-T framework refers to experience, expertise, authoritativeness and trustworthiness. These qualities can be demonstrated through named authors with relevant credentials, first-hand experience, references to primary sources and content that offers something beyond what every other website says.

Original data can make that content more distinctive. Research presented at KDD 2024 on generative engine optimisation found that adding statistics and quotations improved visibility in AI-generated answers by roughly 30 to 40 per cent, while keyword stuffing reduced visibility.

The important point is that adding more keywords wasn't what helped. Evidence, statistics and quotations made the content more useful as a source. Professional firms often have access to material that other businesses would struggle to produce, including anonymised matter outcomes, fee benchmarks, project timelines and insights into the questions clients regularly ask.

Published responsibly, and without disclosing confidential client information, even a small original dataset can give a firm a stronger basis for useful content than another generic list of tips.

The author matters, too. When a partner's name, photograph, credentials and published work are consistent across the firm's website, LinkedIn and industry publications, readers have a clearer way to verify their expertise. This consistency can also help AI systems associate content with a specific expert.

Do Reviews, Mentions, Reddit and YouTube Really Affect AI Recommendations?

Yes, they can. AI systems don't rely exclusively on your website. They may also draw on reviews, directory profiles, industry articles, digital PR coverage and discussions on platforms such as Reddit and YouTube. Consistent coverage in relevant places can help establish a firm's reputation, while a website that only talks about itself offers a narrower picture.

This is an area many firms underinvest in because the work happens outside their own content management system. It includes earning mentions in credible industry publications, maintaining accurate profiles on directories that clients use, encouraging genuine reviews and contributing useful information in places where prospective clients seek advice.

For professional firms, this doesn't mean trying to become an influencer. A partner who answers relevant questions thoughtfully, combined with a handful of clear expert videos on YouTube, can be a sensible starting point. Reviews should be requested appropriately and never purchased.

This outside-the-website work is central to generative engine optimisation. Your website makes the claim, while information elsewhere on the internet helps people and AI systems assess it.

SEO vs AEO vs GEO: What Are You Actually Paying For?

SEO, AEO and GEO address related but different aspects of search visibility. Traditional SEO helps your pages get discovered and ranked. Answer engine optimisation (AEO) makes your content easier to extract and quote. Generative engine optimisation (GEO) focuses on your wider online presence and the evidence that may help AI systems recognise and recommend your firm.

These approaches work best when they support one another. Paying for one while ignoring the others can leave gaps in your visibility where prospective clients are looking for answers.

What Doesn't Work: AI Search Myths That Waste Money

Some of the most expensive mistakes sound technical and create the impression of progress. These include uploading an llms.txt file and waiting for results, stuffing prompts and keywords into pages, buying fake reviews and publishing large volumes of AI-written content. None of these activities, by itself, establishes the clarity, evidence and external credibility that make a firm worth citing.

The llms.txt myth. An llms.txt file is a proposed standard, not an established ranking factor. No major AI system has confirmed that it uses the file to decide which firms to recommend. Adding one may be a small technical experiment, but it should not be sold as a complete strategy.

Keyword and prompt stuffing. The KDD 2024 research cited earlier found that keyword stuffing could reduce AI visibility. Focus on answering the buyer's actual question rather than trying to manipulate a system with repetitive phrases.

“We'll get you into ChatGPT” guarantees. No agency controls ChatGPT's answers. Anyone promising a guaranteed mention is selling certainty they cannot substantiate. A credible programme should explain the work involved, define measurable indicators and report progress honestly over time.

Volume content. Fifty thin, AI-generated posts can leave your website full of repetitive material without giving prospective clients a reason to trust it. A smaller number of useful, original and well-structured pages is a better starting point.

Ignoring Preferred Sources. Google has introduced ways for users to prioritise sources they trust in certain search experiences. Firms with an engaged audience may benefit when readers actively seek out their content. Building a genuine audience is more valuable than trying to manufacture a signal.

Not sure which of these your firm has already paid for? That's a fair question to raise on an initial call. A proper audit should explain what's helping, what's having little effect and what may be wasting your budget.

If you'd like a clearer picture before spending anything else, start with the Gap Report. It shows where your firm's visibility and authority stand against the firms AI is naming instead, and helps identify what to address first.

The AI Search Checklist: 12 Things to Fix, in Order

Start with the basics before moving on to more advanced activity. Establish consistent business information, rewrite your most commercially important pages, implement appropriate schema, publish original data, earn credible external mentions and measure your visibility against named competitors.

1. Write down your firm's core facts. Record its legal name, founding year, locations, services, specialisations and leadership details in one agreed document.
2. Make every profile consistent. Check your website, Google Business Profile, LinkedIn, directories and review platforms. Correct outdated or conflicting information.
3. List the 20 questions buyers actually ask. Use partners' inboxes, intake calls and conversations with clients. Don't rely on keyword tools alone.
4. Rewrite your five most commercially important pages. Put the direct answer first, use question-based headings and include relevant, verifiable details.
5. Add and validate schema. Use appropriate Organisation, Service, FAQPage and Person markup where applicable. Make sure it matches the visible page.
6. Put named experts on your content. Include genuine bylines, credentials, biographies, consistent photographs and links to professional profiles.
7. Publish one piece of first-party data. Consider a benchmark, a survey of your own clients with permission, or an anonymised summary of outcomes.
8. Earn five credible mentions. Prioritise relevant industry publications, professional bodies and reputable directories. Focus on earning coverage rather than buying it.
9. Improve your review process. Ask clients for genuine feedback at an appropriate point in the relationship, using platforms that prospective clients actually consult.
10. Show up where buyers ask their peers. Contribute useful answers to relevant Reddit discussions and consider publishing short expert videos on YouTube.
11. Test your prompts monthly. Use the same 15 to 20 buyer questions in ChatGPT, Google AI Mode and Perplexity. Record which firms appear and how they are described.
12. Track results rather than guessing. Review share of answer, citation rate and AI-referred enquiries monthly. Use the findings to guide quarterly improvements.

How Do You Measure AI Search Visibility?

Measure AI visibility using a fixed set of buyer prompts that you check consistently every month. Record whether your firm is named, where it appears, how it is described and which competitors feature instead. This gives you a basis for tracking citation rate and share of answer, alongside AI-referred website traffic and enquiries.

Avoid overreacting to daily fluctuations. AI answers can vary, so consistency matters more than a single result. Use the same prompts and tools each month, and record the findings honestly.

After three months, you should have a more useful picture of which questions your firm appears for, which questions it misses and which competitors AI systems keep citing. Those gaps can help guide your next round of content and authority improvements.

Many firms incorporate this work into a broader search everywhere optimisation programme. Prospective clients don't necessarily separate their behaviour into Google users and ChatGPT users, so measurement should reflect how they search across different platforms.

Should You Hire Help for AEO and GEO, and What Should It Include?

Hiring help makes sense when the time required from your partners outweighs the cost of bringing in specialists. Many firms struggle to complete entity clean-up and substantial content rewrites alongside their day-to-day client work.

A proper AEO and GEO engagement should include an entity audit, answer-first rewrites of commercially important pages, schema implementation, a plan for external mentions and digital PR, and monthly reporting on share of answer against named competitors.

If a proposal consists mainly of a tool subscription and a content calendar, ask what strategic work is actually being delivered.

Before hiring an agency, ask three questions:

1. Which exact buyer prompts will you track before work begins?
2. Which pages will you rewrite first, and why have you prioritised them?
3. What measurable evidence will show whether the programme is working after 90 days?

Clear answers suggest that the agency has a defined process. Vague promises make it difficult to judge whether the retainer is delivering value.

Frequently Asked Questions

How Long Does It Take to Get Cited in AI Search?

Structural improvements, including consistent business information, schema and answer-first rewrites, may begin to appear in AI answers within weeks as systems recrawl and reassess sources. Meaningful changes in recommendations can take three to six months or longer as external mentions, reviews and original content accumulate. Results vary, and no agency can guarantee a specific recommendation by a fixed date.

Is AEO Replacing SEO?

No. AI answers draw on pages and sources that traditional SEO helps make discoverable. AEO and GEO extend that work by focusing on extraction, citation and recommendation. Firms that abandon their SEO foundations in favour of AEO risk weakening both their conventional rankings and their AI visibility.

What Is AI Search Visibility, in Plain Terms?

AI search visibility describes how often, and how favourably, AI systems mention your firm when prospective clients ask relevant questions. You can measure it by tracking citation rate and share of answer across a consistent set of prompts. The aim is not simply to appear in a list, but to be included in the recommendations that matter to your business.

Does My Firm Need to Be on Reddit and YouTube?

You should consider the platforms where your prospective clients seek information and peer opinions. Reddit discussions and YouTube explainers can contribute to your wider online presence. You don't need to become an influencer. Helpful contributions from partners and a small collection of clear, informative videos can be a reasonable starting point.

What Does an llms.txt File Actually Do?

It is a proposed standard intended to give AI crawlers a structured overview of a website. Major AI providers have not confirmed that it influences recommendations. It may be worth experimenting with if implementation is straightforward, but it should not take priority over useful content, consistent business information and credible external references.

Can a Small Firm Beat a Big Firm in AI Answers?

Yes, it is possible. A focused firm with a clear specialisation, consistent business information, strong reviews and original material may be easier to describe accurately than a larger firm with a vague or inconsistent online presence. A well-defined niche can be an advantage, although no particular outcome is guaranteed.

Your Buyers Are Already Asking AI Who to Hire

Somewhere this week, a founder, general counsel or family business owner asked an AI tool to recommend a firm like yours and received three names. The question is whether your firm was one of them. You don't have to guess.

Start with the Gap Report. We'll show you where your firm stands against the firms AI is naming, which of the 12 checklist items deserves attention first and what a realistic 90-day plan could look like.

There are no guarantees about ChatGPT recommendations, just a clearer picture of your current position and the work needed to improve it.