Introduction – the shift from “10 blue links” to agentic search
Search engines and browsers are becoming agentic – they read content, summarize it and perform actions on behalf of users. Google’s Search Generative Experience (SGE) and AI overviews give direct answers at the top of the results page, significantly reducing the need to click through. Studies across SEO tools show that AI overviews reduce position‑1 click‑through rates (CTR) by 34–40 %; Amsive found a 15.49 % decline and Ahrefs reported a 34.5 % dropomnius.so. BrightEdge data indicates that while search impressions rose by 49 % in 2024–2025, click‑through rates dropped 30 %omnius.so.
The adoption of AI overviews is accelerating; they were triggered for ~6.49 % of queries in January 2025 and climbed to 13.14 % by March 2025 (72 % monthly growth)omnius.so. By mid‑2025 the presence of AI overviews halved paid‑ad CTRsdinocajic.com and some publishers reported double‑digit traffic dropsdinocajic.com. WordStream’s August 2025 statistics show that AI overviews appear in ~55 % of Google searches and have grown 115 % since March 2025wordstream.com. They take up 42 % of desktop screens and nearly half of mobile screens, pushing organic results far downwordstream.com. Users mostly read the first third of an AI overview, with 7 in 10 searchers only scanning the first few lineswordstream.com.
These changes are altering the competitive landscape. ChatGPT and other AI search tools now command a noticeable share of queries; for instance, one industry article notes that ChatGPT handled 80.1 % of the AI‑search market and 9 % of global queries in 2025 while Google’s share dipped below 90 % for the first time since 2015ampcome.com. Instead of providing links, AI agents perform multi‑step tasks: they compare options, fill forms and deliver solutionsampcome.com. For marketers, this means optimizing content for machine understanding and building actions into pages rather than relying on ranking alone.
The rise of vertical AI and agentic systems
From horizontal SaaS to vertical AI agents
General‑purpose large‑language models (LLMs) are giving way to vertical AI agents – domain‑specific systems that combine reasoning engines, domain training and real‑time adaptability. A 2025 Turing report notes that general‑purpose SaaS is being supplanted by vertical AI agents that embed deep domain expertise into workflowsturing.com. Bessemer Venture Partners projects that vertical AI market capitalization could grow ten‑fold compared with legacy SaaS and estimates the market will surpass US $100 billion by 2032turing.com. Turing cites a forecast that 80 % of enterprises will adopt vertical AI by 2026turing.com.
The shift is driven by the shortcomings of horizontal SaaS platforms. Turing notes that traditional SaaS lacks contextual intelligence, relies on static workflows and imposes heavy customization burdensturing.com. Vertical AI agents address these issues by integrating domain‑specific reasoning, real‑time adaptability and compliance alignmentturing.com. They act as expert operators within an industry – turning doctor‑patient conversations into notes (Abridge), generating demand letters (EvenUp) or predicting manufacturing equipment failures (Axion Ray)turing.com. Bessemer observes that vertical AI adoption has accelerated even among “technophobic” verticals, because these systems solve language‑heavy workflows that previous software couldn’tbvp.com.
Vertical AI as a new control point
Venture firm NEA argues that the control point in enterprise software is shifting from systems of record (CRM/ERP) to agentic systems of action. In the past, vertical software companies built defensibility through data ownership and high switching costs; AI agents now process data before it reaches those systemsnea.com. NEA predicts that agentic systems will unlock the US$11 trillion U.S. labour spend—vastly larger than the US$450 billion enterprise software market—by automating labour‑heavy workflowsnea.com. The cost of machine intelligence is declining due to intense competition among foundation‑model providers, making it feasible to build industry‑specific agentsnea.com.
Bessemer identifies clear patterns among breakout vertical AI companies: they start with a compelling wedge (often language‑heavy or multi‑modal tasks), embed deeply into existing workflows, build defensible data moats, and deliver immediate ROI—reallocating labour to higher‑value work or driving topline growthbvp.com. However, the report highlights open questions: whether vertical AI companies will integrate with or replace legacy systems of record; how they will compete with incumbents; and whether they can maintain data advantages in privacy‑sensitive industriesbvp.com.
Cross‑industry adoption statistics
Overall AI adoption and market size
The global AI market is expanding rapidly. Coherent Solutions reports that AI adoption grew from 55 % to 75 % between 2023 and 2024, delivering a 3.7× return on investment for adopterscoherentsolutions.com. Goldman Sachs projects that AI could boost global GDP by 15 % over the next decadecoherentsolutions.com, though more conservative estimates peg it at 8 %–9 %coherentsolutions.com or even 1–1.5 %. Within the top 25 % of AI spenders are healthcare, financial services, media & telecom, manufacturing and retailcoherentsolutions.com, followed by energy, consumer goods, hardware engineering, travel/transport and logisticscoherentsolutions.com.
Healthcare
Healthcare is becoming a pace‑setter for enterprise AI adoption. Menlo Ventures notes that the $4.9 trillion U.S. healthcare industry, which historically lagged in digital transformation, is now deploying AI at 2.2× the rate of the broader economymenlovc.com. Adoption jumped from 3 % in 2023 to 27 % of health systems, 18 % of outpatient providers and 14 % of insurers by 2025menlovc.com. Twenty‑two percent of healthcare organizations have implemented domain‑specific AI tools, representing a 7× increase over 2024menlovc.com. AI spending in healthcare reached US $1.4 billion in 2025—nearly triple the 2024 amountmenlovc.com—and the sector produced eight AI unicornsmenlovc.com. Kaiser Permanente deployed Abridge’s ambient documentation system across 40 hospitals and 600+ medical offices, the largest generative AI rollout in healthcare historymenlovc.com. Advocate Health evaluated over 225 AI solutions and selected 40 to go live, projecting documentation time reductions of more than 50 %menlovc.com.
Legal services
Legal AI adoption has accelerated but remains uneven. A 2025 legal industry report highlighted in Cicerai’s survey found that 21 % of law firms already use generative AI, and nearly one‑third more plan to adopt it by year endcicerai.com. The explosion of digital evidence and rising client expectations are driving demand for AI tools that automate document drafting, research, contract lifecycle management and e‑discoverycicerai.com. However, other surveys (outside this citation scope) still indicate that large portions of law firms have not yet implemented AI, suggesting significant growth potential.
Finance and banking
AI is driving productivity in finance by automating compliance, anomaly detection and robo‑advisory. Coherent Solutions notes that the financial sector could gain US $1.2 trillion in additional gross value added (GVA) by 2035 due to mass AI adoptioncoherentsolutions.com. AI tools are already used for payments, fraud detection and algorithmic tradingcoherentsolutions.com. However, governance by central banks and regulatory risks may slow adoptioncoherentsolutions.com.
Manufacturing and supply chain
Manufacturers are enthusiastic adopters of AI. The 2025 State of AI in Manufacturing survey found that over 77 % of manufacturers have implemented AI (up from 70 % in 2023)coherentsolutions.com. Adoption is highest in production (31 %), customer service (28 %) and inventory management (28 %)coherentsolutions.com. Manufacturers are most interested in collaborative “copilot” agents that augment human workers—53 % prefer collaborative bots over full autonomycoherentsolutions.com. Investment is concentrated in supply chain management (49 %) and big‑data analytics (43 %)coherentsolutions.com.
Retail and e‑commerce
Retailers are using generative AI to personalise marketing and improve conversions. Deloitte’s 2025 US retail industry outlook noted that retailers saw 15 % higher conversion rates during Black Friday when chatbots assisted customerscoherentsolutions.com. IBM research suggests that organizations in retail and consumer products will make the most extensive use of AI across 2025 and beyondcoherentsolutions.com. Use cases include conversational agents for customer service, demand forecasting, pricing optimisation and generative product descriptions.
Media, publishing & SEO‑driven websites
Media publishers are among the biggest losers in the agentic search era. AI overviews reduce organic CTR drastically; one study of ~10,000 informational queries found that when an AI overview appears, overall organic CTR drops from 1.41 % to 0.64 %dinocajic.com. Another analysis showed organic clicks shrinking by 18 % to 64 % for publishersdinocajic.com. Specific sites have reported severe traffic declines; for example, HubSpot’s marketing blog saw an estimated 80 % plunge in organic traffic within a year after AI overviews rolled outdinocajic.com. Paid advertising suffers too: Seer Interactive found that when AI overviews are present, paid‑ad CTRs nearly halved from ~21 % to 10 %dinocajic.com.
Telecom & IT services
Telecom providers use AI for network planning, optimization, security, customer‑experience enhancement and predictive maintenancecoherentsolutions.com. The AI‑RAN Alliance launched in February 2024 to integrate AI with radio‑access networkscoherentsolutions.com. Deloitte estimates that AI could add US $4.7 trillion in GVA to IT and telecom by 2035coherentsolutions.com.
Education
AI tools tailored for teachers are emerging. Bessemer highlights companies like Brisk Teaching and MagicSchool, which automate grading, tutoring and content creationbvp.com. These tools illustrate how vertical AI can support educators by handling routine tasks, freeing teachers to focus on pedagogy and student engagement. Adoption remains nascent but is growing as generative AI becomes integrated into learning management systems.
Real estate & home services
Vertical AI companies such as EliseAI automate property management workflows—from prospect communications to lease audits—while Hatch acts as AI‑powered customer service for home‑service providersbvp.com. These agents operate as virtual assistants, reducing operational costs for landlords and home‑service businesses. New entrants like Rilla analyse in‑person sales conversations to coach repsbvp.com. Adoption is early but demonstrates how language‑heavy interactions can be automated.
Competition dynamics and market implications
Reallocating traffic and value
The rise of agentic search is shrinking traffic for traditional websites. AI overviews provide direct answers, leading to fewer clicks and more zero‑click searches—already nearly 60 % of Google queries in 2024dinocajic.com. When an AI answer is expanded, the top organic result can be pushed down by ~1,500 pixels – more than a full screendinocajic.com. Users often trust the AI answer: 70 % say they somewhat trust generative search resultswordstream.com and 79 % expect to use AI‑enhanced search within a yearwordstream.com. As AI overviews appear in over half of searches and prioritise high‑authority sourceswordstream.com, smaller sites may see their visibility vanish unless their content is cited or embedded in AI answers.
Vertical AI start‑ups vs incumbents
The explosion of vertical AI has created new competitive fronts. Bessemer notes that vertical AI companies are solving high‑value, language‑heavy tasks that incumbent SaaS platforms neglectedbvp.com. Many incumbent software vendors are therefore buying vertical AI start‑ups or building their own agentic features. Bessemer lists examples across healthcare (Abridge, SmarterDx), legal (EvenUp, Ivo, Legora), education (Brisk Teaching, MagicSchool), real estate (EliseAI), home services (Hatch) and othersbvp.com. Success requires domain expertise, context‑rich data and seamless workflow integrationbvp.com.
NEA warns that the control point is shifting: AI agents operate upstream of traditional systems of record, potentially eroding the incumbents’ data moats and switching costsnea.com. Incumbent players risk being relegated to commodity data stores if they fail to build agentic layers. On the other hand, vertical AI start‑ups must defend their data advantages; as they expand into new tasks, regulatory barriers and fragmented datasets may challenge their competitivenessbvp.com.
Regulatory and ethical considerations
As AI agents take on complex tasks—drafting medical notes, generating legal documents or managing financial transactions—regulatory scrutiny increases. Healthcare agents must comply with HIPAA and medical device regulations; financial agents face Basel III and anti‑money‑laundering rules; legal agents must meet confidentiality and professional‑ethics standardsturing.com. Data privacy laws (GDPR, CCPA) may limit the data agents can use, while copyright questions arise when AI outputs derive from training data. Firms building vertical agents need robust compliance frameworks and transparent auditing to avoid legal risks.
Impact on advertising and revenue models
With fewer clicks, advertising revenue tied to traditional pay‑per‑click models may decline. Paid search CTRs drop by around 50 % when AI overviews are presentdinocajic.com, and the AI answer often leapfrogs paid placementsdinocajic.com. This means advertisers may need to shift budgets to AI‑powered recommendation engines, sponsored answers, or embedded commerce within AI agents. Publishers reliant on display ads must experiment with other revenue streams such as subscriptions, premium tools and data licensing (e.g., licensing content for AI training or quoting).
How verticals can benefit and adapt
To thrive in the agentic era, businesses need to satisfy both humans and machines. The following actions—drawn from the evolving SEO/agentic guidelines—can help capture traffic and harness vertical AI:
- Provide first‑click fulfilment (FCF) – Offer a concise summary at the top of each page answering common questions, including price ranges and clear calls to action. Present machine‑friendly FAQs and mark them up with
FAQPageandSpeakableSpecificationschema. When AI agents crawl pages, they will extract and cite these summaries, increasing the likelihood of being referenced in AI overviews. - Expose structured data and machine‑readable information – Use schema.org vocabularies (e.g.,
Product,Review,HowTo,ItemList) to label key details. For vertical sites, include domain‑specific attributes (e.g., medical procedure codes, legal case types, financial instruments) so agents can understand and compare offerings. AddOfferorAggregateOffermarkup to affiliate links and pricing to make them discoverable. - Embed calculators and interactive tools – AI agents will prefer pages that perform actions. Build small utilities (e.g., dosage calculators, mortgage payment estimators, fan CFM calculators) with inputs and outputs exposed in HTML. Support URL parameters (
?prefill=) so AI agents can deep‑link directly to computed results. - Publish authoritative, in‑depth content – AI overviews prioritize high‑authority sources and domain experts. Investing in expert authorship, citations and transparent methodologies will increase the chances of being cited. For healthcare content, collaborate with clinicians; for legal content, reference statutes; for finance, include regulatory guidance.
- Develop vertical AI agents or partner with specialists – Enterprises should assess which workflows can be automated and either build or adopt vertical agents. For instance, healthcare providers may deploy ambient scribe agents like Abridge; law firms might use Harvey for contract analysis; manufacturers could adopt predictive‑maintenance agents. Early adopters gain productivity gains and data‑moat advantages.
- Monitor and optimize for AI overviews – Track which queries trigger AI overviews in your vertical and adjust strategies accordingly. WordStream reports that AI overviews show up in only 7 % of local querieswordstream.com; local businesses can continue focusing on local SEO. For informational queries where AI overviews dominate, focus on being cited within the AI answer rather than ranking first.
- Prepare for new monetization models – Explore subscription or membership models, offer premium tools (e.g., detailed analytics, personalized recommendations) and consider licensing data/models to AI platforms.
Vertical‑by‑vertical summary table
| Vertical | Adoption & market trends (2025) | Key benefits/opportunities | Competitive implications |
|---|---|---|---|
| Healthcare | Adoption surged to 27 % of health systems, 18 % of outpatient providers and 14 % of insurersmenlovc.com; AI spending reached US $1.4 B (3× 2024 levels)menlovc.com; Abridge deployed across 40 hospitals & 600+ medical officesmenlovc.com. | Automate clinical documentation, reduce administrative burden, accelerate drug discovery; ambient scribe agents improve clinician productivity and reduce burnout; predictive analytics for resource allocation. | New entrants like Abridge, SmarterDx and OpenEvidence challenge EHR vendors; incumbents must integrate AI or risk becoming commodity data stores; regulatory compliance (HIPAA, FDA) critical.bvp.com |
| Legal services | 21 % of law firms use generative AI and one‑third plan adoptioncicerai.com; AI tools automate drafting, research and e‑discoverycicerai.com. | Agents like Harvey and EvenUp generate demand letters, review contracts, perform legal research; improve efficiency, reduce billable hours and expand access to justice. | Fragmented adoption creates opportunity for vertical AI vendors; incumbents may license AI or acquire start‑ups. Ethical oversight and client confidentiality are major concerns. |
| Finance & banking | Sector could gain US $1.2 T extra GVA by 2035 through AIcoherentsolutions.com; uses include payments, fraud detection, robo‑advisorscoherentsolutions.com. | AI agents streamline onboarding, risk assessment, compliance (AML/KYC), algorithmic trading; personal finance agents democratize advice. | Regulatory scrutiny (Basel III, GDPR) may slow adoption; incumbents need robust governance; competition from FinTechs and Big Tech vertical agents intensifies. |
| Manufacturing & supply chain | >77 % of manufacturers have implemented AIcoherentsolutions.com; major adoption in production (31 %), customer service (28 %) and inventory management (28 %)coherentsolutions.com; focus on supply chain (49 %) and analytics (43 %)coherentsolutions.com. | Predictive maintenance, demand forecasting, quality control, autonomous planning; collaborative robots (cobots) augment workerscoherentsolutions.com. | Vertical AI start‑ups like Axion Ray compete with industrial IoT vendors; integration with legacy ERP is a challenge (56 % unsure of readinesscoherentsolutions.com). |
| Retail & e‑commerce | AI chatbots improved conversion rates by 15 % during Black Fridaycoherentsolutions.com; sector expected to make extensive use of AIcoherentsolutions.com. | Personalized recommendations, dynamic pricing, demand forecasting, image‑based search; agents handle product inquiries and returns. | AI overviews may cannibalize traffic to product pages; retailers must ensure their data feeds (pricing, availability) are machine‑readable; competition from AI‑native shopping assistants intensifies. |
| Media & publishing | Organic CTR drops from 1.41 % to 0.64 % when AI overviews appeardinocajic.com; publishers report 18–64 % traffic dropsdinocajic.com; HubSpot saw an 80 % declinedinocajic.com. | Opportunity to license content to AI models, develop interactive tools and subscription products; emphasise authoritative reporting to be cited in overviews. | Advertising revenues decline as AI answers displace clicks; large platforms with strong authority benefit, while smaller sites struggle; new revenue models required. |
| Telecom & IT services | AI could add US $4.7 T GVA by 2035coherentsolutions.com; AI‑RAN Alliance launched to merge AI with cellular technologycoherentsolutions.com. | Network optimization, predictive maintenance, customer‑service bots, network slicing; vertical AI agents manage spectrum allocation and service provisioning. | Telecom vendors must embed AI to remain competitive; open initiatives (AI‑RAN) may level the field, but data privacy and infrastructure costs pose challenges. |
| Education | Early adoption of teacher‑focused AI tools (Brisk Teaching, MagicSchool)bvp.com. | Automate grading, tutoring, lesson planning; personalize learning; voice‑based tutoring. | Ed‑tech start‑ups and big‑tech providers compete; privacy (FERPA) and bias concerns must be addressed. |
| Real estate & home services | Agents like EliseAI automate property management; Hatch provides AI customer servicebvp.com. | Automate tenant communications, leasing, maintenance; voice/speech analytics for sales coachingbvp.com. | Traditional property‑management software may be displaced; adoption depends on integration with existing systems and compliance with housing regulations. |
Conclusion
Agentic search and vertical AI represent a fundamental reshaping of the digital economy. AI overviews and generative search are reducing clicks and forcing websites to deliver immediate answers and machine‑readable data. At the same time, vertical AI agents are automating domain‑specific workflows, promising productivity gains across healthcare, law, finance, manufacturing, retail and more. Adoption is advancing fastest in industries with abundant structured and unstructured data (healthcare, finance, manufacturing) but is spreading to traditionally “technophobic” sectors such as education, real estate and home services.
For businesses, surviving and thriving in this environment requires designing for agents: structuring content with schema, providing first‑click fulfilment, embedding interactive tools, developing or partnering with vertical AI agents and exploring new monetization models. Companies that embrace these changes early will harness the efficiencies of automation and maintain visibility in a world where AI agents increasingly control the flow of information and customer interactions.
