Written by: Arjun Karnik, Growth Marketing Specialist
Key Takeaways
- Generative engine optimization (GEO) helps professional services firms get cited and recommended in AI answer engines like ChatGPT, Google AI Overviews, Perplexity, and Gemini.
- Buyer behavior has shifted, with 68% of Google searches ending without a click and 51% of B2B buyers now starting their journey on AI chatbots.
- Arjun Karnik’s seven-step GEO system covers visibility audits, technical plumbing, fan-out query mapping, buyer-language alignment, structured publishing, freshness loops, and citation measurement.
- Success metrics have moved from traditional rankings to AI citations, share of answer, and AI referrer traffic, with early citations compounding into durable advantage.
- See how these visibility shifts affect your firm’s practice areas by booking a demo with Arjun Karnik.
Why professional services firms need GEO now
The buyer behavior shift already shows up inside firms that have followed SEO best practices for years. SparkToro’s analysis of Similarweb clickstream data for January through April 2026 found that 68.01% of U.S. Google searches ended without a click, up from 60.45% in 2024. When an AI Overview appears, only 8% of users click a traditional result versus 15% without one, which is a 47% reduction in click-through.
The legal vertical shows the impact most clearly. A non-Semrush source reported that 78% of legal queries trigger Google AI Overviews, while Semrush’s analysis of 10M+ keywords did not break out this percentage for legal queries. AI referral traffic to UK law firm sites grew 8.4x between January 2025 and June 2026, which reflects a structural channel shift, not a short-term spike.
A March 2026 G2 survey of 1,076 B2B software buyers found that 51% now start their purchase journey on an AI chatbot rather than a traditional search engine, up from 29% eleven months earlier. Sixty-nine percent changed their intended vendor based on what the assistant told them, and 33% bought from a vendor they had never previously heard of. Being in the answer functions as a vendor-selection event, not a soft visibility metric.
At Google I/O in May 2026, Sundar Pichai reported AI Overviews at over 2.5 billion monthly active users and AI Mode at over 1 billion monthly active users within its first year. This channel no longer runs as a side experiment. It already behaves as the primary search interface many buyers use every day.
Firms benefit most when they start GEO before answers harden and citation patterns stabilize. Early citations become tomorrow’s record, and AI Growth Agent clients average more than 12,000 additional AI citations and mentions and a 20% or greater lift in impressions across the first twelve weeks.
The seven GEO strategies in Arjun Karnik’s system
- Visibility audit. Baseline current citations across ChatGPT, Google AI Overviews, Perplexity, and Gemini before publishing anything new. Treat this as the control group that every later result is measured against.
- Technical plumbing. Unblock AI crawlers, add schema markup, and make pages machine-parseable. Without this foundation, every downstream investment lands on content the retrieval layer cannot reliably access.
- Fan-out query mapping. Extract the full question space behind a buyer prompt directly from ChatGPT instead of inferring it from keyword tools. On Arjun’s own site, pages rewritten to match extracted fan-out queries earned citations while control pages did not.
- Buyer-language alignment. Rewrite slugs, titles, H1s, and H2s to match the words buyers use, not the words practitioners prefer. On Arjun’s own site, relabelling a jargon-heavy page to buyer language produced citations within weeks of that specific change.
- Structured publishing at machine cadence. Deploy an AI article engine on a site subfolder and publish at 5 to 8 autonomous actions per day via AI Growth Agent, combining new articles with updates. On Arjun’s own site, the GEO subfolder went from zero to the only source of new impressions on the domain in 60 days.
- Freshness loop and self-healing content. Set impression-decay tripwires that auto-queue updates when performance drops. In Arjun’s own tests, pages dropped 78% to 99% in two months without maintenance. Approximately 50% of sources cited for a given prompt change within 13 weeks, so a fixed library of any size decays in place.
- Citation and share-of-answer measurement. Track citations across all four surfaces, AI referrers such as chatgpt.com in analytics, and impression and decay curves in Google Search Console. Share of answer replaces rank position as the primary success metric.
How fan-out queries and topical authority drive share of answer
A single buyer prompt rarely produces a single lookup. It triggers dozens of hidden retrieval queries underneath, and the answer is assembled from those results. Legacy tools like Semrush and Ahrefs rely on historical data and daily caps, which miss new long-tail queries that AI surfaces answer. Focusing only on the visible prompt while ignoring the fan-out means targeting the wrong surface.
Topical authority in this channel grows through depth of coverage across a mapped question space, not through inherited backlink strength. Adding statistics increases AI citation visibility by around 31–33% and adding quotations by around 41–43%, according to the Princeton GEO study. A competitive share of citation for B2B brands in 2026 sits between 5% and 15% aggregate across major AI engines, with 20% or above signaling category leadership.
Share of answer is calculated by running a fixed set of 50–100 target queries across ChatGPT, Gemini, Perplexity, and Claude, then measuring how often the brand appears as a cited source. Brands earning both a mention and a citation in AI-generated answers are up to 40% more likely to maintain ongoing visibility, which makes this combined footprint a practical benchmark.
Structured data requirements for GEO-ready sites
Structure acts as a requirement rather than a finish line because it forms the foundation that makes every other GEO activity work. B2B sites with comprehensive structured data can achieve higher citation rates because AI engines can parse and cite their content more reliably. The five-tier schema stack for professional services firms runs from entity foundation (Organization, Person, WebSite) through content context (Article, BlogPosting, BreadcrumbList), answer surfaces (FAQPage, HowTo), commercial layer (Service), and trust signals (Review, AggregateRating) only when those assets genuinely exist.
Ahrefs 2026 data shows only 38% of AI Overview citations now come from organic top-10 pages, down from 76% a year earlier, which confirms that schema-supported entity authority increasingly determines citation selection over traditional ranking position. For consulting and professional services firms, Person schema anchored to stable @id URIs with worksFor linking to Organization, knowsAbout expertise areas, and sameAs links to LinkedIn and Wikidata carries disproportionate weight because AI engines cite named expertise over brand-level claims.
Domain authority shows only a +0.18 correlation with AI citation rates. Technical plumbing such as unblocked AI crawlers, schema coverage, and machine-parseable pages must be fixed before any content strategy, because nothing downstream functions without this layer.
Step-by-step GEO implementation for firms
The implementation sequence follows the seven-step system with concrete actions at each stage.
- Run the visibility audit. Query ChatGPT, Gemini, Perplexity, and Google AI Overviews with 50–100 buyer-intent prompts. Log where the firm appears, where competitors appear instead, and where no structured answer exists.
- Fix technical plumbing. Confirm AI crawlers are unblocked in robots.txt. Deploy Organization and Person schema in JSON-LD with sameAs links to LinkedIn, Wikidata, and relevant professional registries. Validate in Google’s Rich Results Test before publishing.
- Map fan-out queries. Extract the sub-queries ChatGPT generates underneath a buyer prompt. Use these as the production queue that defines what gets written and in what language.
- Rewrite to buyer language. Align slugs, titles, H1s, and H2s to the extracted question language. Make the first sentence under each H2 fully answer the heading so retrieval systems can extract the passage cleanly.
- Publish at machine cadence via AI Growth Agent. Deploy a structured article engine on a site subfolder running 5 to 8 autonomous actions per day. On Arjun’s own site, new articles reached thousands of monthly Google impressions within weeks.
- Set impression-decay tripwires. Wire tripwires to Search Console signals so updates queue automatically when performance drops. Recently updated pages can receive more AI citations.
- Monitor citations and AI referrers. Track share of answer monthly. Segment chatgpt.com and equivalent referrers in analytics as a distinct traffic class and feed wins back into the production queue.
SEO vs GEO: which metrics actually matter now?
| Dimension | Traditional SEO | Generative Engine Optimization (GEO) |
|---|---|---|
| Optimizes for | Human-ranked lists and domain authority | Machine retrieval and citation in AI answers |
| Query model | The keyword the buyer typed (average 3.37 words) | Dozens of hidden fan-out queries triggered by one prompt (average ChatGPT prompt is 23 words) |
| Primary success metric | Rank position | Citations, share of answer, AI referrer traffic |
| Where authority comes from | Backlinks and domain authority | Expert topical coverage and entity signals |
| What sustains a win | Accumulated domain authority | Continuous freshness supported by frequent updates |
Why GEO metrics replace rank as the headline KPI
GEO replaces a single rank report with a three-layer stack that covers visibility, engagement, and business outcomes. Visibility KPIs act as leading indicators, engagement KPIs form the bridge, and business KPIs lag but confirm revenue impact.
The core visibility metrics are:
- AI Answer Presence Rate: the percentage of relevant queries for which the firm appears in an AI-generated response. A practical goal is presence in 30–40% of relevant niche queries within the first six months of an active GEO strategy.
- Share of Citation: calculated as (citations to your domain ÷ total citations in the tracked prompt set) × 100. AI citations change 40–60% monthly, which makes monthly monitoring the minimum viable cadence.
- AI referrer traffic: sessions arriving from chatgpt.com and equivalents, segmented in analytics. Rankability and Microsoft Clarity data indicate AI-referred traffic converts to sign-ups at about 1.66% versus 0.15% for organic search.
- Impression and decay curves in Google Search Console: the scissors chart, where impressions rise while clicks fall, signals that content is being consumed by AI systems without sending traffic back.
One caveat applies to every figure in this framework. Buyers frequently copy an answer and type a name directly into a browser, which lands in analytics as direct or branded search. Whatever is measured represents a floor, not a ceiling.
Common GEO objections and pitfalls for firms
The most common objection claims that GEO simply renames SEO. The target has changed in practice. SEO focuses on rankings on a human-readable list, while GEO focuses on citation inside a machine-generated answer. Retrieval mechanics, success metrics, and authority models differ across the two systems.
The most common pitfall treats freshness as hygiene rather than the core game. As mentioned in the freshness loop strategy, content decay happens quickly, so treating updates as optional maintenance creates a structural disadvantage. Citation half-life for AI sources is approximately 3.4 weeks on ChatGPT and 4.3–4.8 weeks on Google AI surfaces. A quarterly audit cycle moves too slowly for a game that resets weekly.
A second pitfall involves publishing without structure. AI-cited articles often include more statistics and expert quotes than non-cited articles, which aligns with the Princeton GEO findings. Prose that is not structured for retrieval, not aligned to fan-out query language, and not carrying schema will lose to a weaker page that satisfies all three requirements.
A third pitfall involves waiting for the channel to mature. AI platforms cite a limited number of brands per response, and a small group of domains already accounts for a large share of AI citations. Early citations become tomorrow’s record, and the cost of entry rises as settled answers harden.
A fourth pitfall ignores what AI already says about the firm. A wrong AI answer hurts more than no answer. Defensive GEO, which audits and corrects what assistants currently say about the firm, runs in parallel with growth work rather than after it.
FAQ: GEO for law, accounting, and other professional services
What is the difference between GEO and traditional SEO for a law firm or accounting practice?
Traditional SEO earns rankings on a list of ten blue links, while GEO earns citations inside a machine-generated answer. SEO optimizes against the keyword the buyer typed, and GEO optimizes against dozens of fan-out sub-queries the buyer never sees. Authority in SEO accumulates through backlinks and domain age, while authority in GEO is built through topical coverage depth, entity signals, and continuous freshness. A law firm that ranks on page one can still be absent from AI answers if its content is unstructured, stale, or misaligned to buyer-language questions.
How long does it take to see citations in AI answers after starting GEO?
Coverage and impressions typically appear within weeks, and citations in AI answers follow in one to three months. Compounding, where topical authority accumulates and share of answer grows, usually begins after month three. On Arjun’s own site, new articles reached thousands of monthly Google impressions within weeks of publication, and the GEO subfolder became the only source of new impressions on the domain in 60 days. These figures reflect Arjun’s own test-lab results, not guaranteed outcomes for any other property.
How do you measure GEO results when Google Search Console does not report AI citations?
The measurement system uses four instruments in parallel. First, run a fixed set of 50–100 buyer-intent prompts across ChatGPT, Gemini, Perplexity, and Claude monthly and log every citation. Second, segment AI referrers such as chatgpt.com as a distinct traffic class in analytics, because they convert like referrals rather than cold search traffic. Third, track impression and decay curves in Search Console to catch the scissors pattern before it becomes a revenue problem. Fourth, monitor branded search volume for spikes that follow AI visibility gains, since many buyers copy an answer and type a name directly into a browser rather than clicking a citation link. Whatever is measured represents a floor, not a ceiling.
Does GEO require stopping traditional SEO?
GEO does not require abandoning traditional SEO. Technical fundamentals, structured content, and quality writing support both channels. What changes involves the optimization target and the reporting metric. Content built for AI citation still performs in traditional Google search. On Arjun’s own site, articles structured for GEO reached thousands of monthly Google impressions within weeks. The practical shift moves from rank tracking as the headline KPI to share of answer and citation frequency, and adds the freshness loop that traditional SEO retainers rarely include.
What does the self-verifying nature of Arjun Karnik’s test lab mean in practice?
The test lab is self-verifying because the same system being documented produces the visibility. Asking an AI assistant about generative engine optimization topics reveals who gets cited. The method functions as the proof. Arjun publishes specific tests, numbers, and misses from his own site, not case studies written after the fact, but the actionable work itself, including what was published, what was restructured, what was refreshed, and what happened. His figures come from his own Google Search Console, cadence records, and decay curves. AI Growth Agent’s case studies are cited separately as AI Growth Agent’s results and are never blended with his own data, so the receipts remain public and checkable without relying on marketing claims.
Ready to get your firm cited in AI answers?
Professional services firms that rank in traditional search but remain absent from AI answers lose vendor-selection events before a buyer ever visits their site. The seven-step GEO system, which covers visibility audits, technical plumbing, fan-out query mapping, buyer-language alignment, structured publishing at machine cadence via AI Growth Agent, freshness loops, and citation measurement, provides a repeatable and self-verifying approach that Arjun Karnik runs in public and documents with receipts.
