Google's Business Agent for Leads, announced at Google Marketing Live on May 20, 2026, replaces the static lead form inside Search ads with a Gemini-powered chat agent that answers prospect questions in real time using the advertiser's website content. Instead of clicking through to a landing page and filling out a form, prospects engage in a conversation directly within the ad, and the system submits a pre-filled lead form only when the exchange indicates genuine intent. The format launched in open beta for US advertisers in automotive, education, and real estate verticals, running exclusively through AI Max for Search or Performance Max campaigns.
What Google Announced at Marketing Live 2026
Google held its annual Marketing Live event on May 20, 2026, as a free virtual gathering focused on turning the company's "Gemini advantage" into practical advertising tools. The keynote introduced what Google called a new generation of ads for the AI era of Search, a suite of conversational, contextual ad formats designed to make advertising feel less like interruption and more like assistance.
Business Agent for Leads was one of four formats announced:
- Conversational Discovery ads, which answer a person's specific question directly inside AI Mode using Gemini-generated creative tailored to the query
- Highlighted Answers, which surface relevant ads as recommendations when AI Mode generates a list (such as language apps for an upcoming trip)
- AI-powered Shopping ads, expected to roll out later in 2026 for US advertisers
- Business Agent for Leads, the B2B and lead-generation pillar of this push
How Business Agent for Leads Works
The format replaces the traditional lead form with a chat interface. When a prospect clicks a Search ad, they see a "Chat" button instead of a form. Clicking it opens a conversation window where the prospect can ask questions about the product, service, pricing, availability, or process. The agent answers in real time, drawing responses from the advertiser's website content.
The prospect never visits the landing page during this exchange. The conversation happens inside the ad unit. Once the dialogue signals real buying intent (the prospect has asked meaningful questions, received answers, and expressed interest), the system pre-fills a lead form with the contact information it has captured and submits it to the advertiser.
The advertiser receives not just a name, email, and phone number, but the conversational context: which questions the prospect asked, what concerns they raised, what priorities they expressed, and which features or details mattered most to them. This gives sales teams richer qualification data than a traditional form submission, where the only signal is that the prospect filled out fields.
| Element | Traditional Lead Form | Business Agent for Leads |
|---|---|---|
| User Action | Click ad, land on page, fill form | Click "Chat", ask questions in ad |
| Content Source | Static landing page copy | Gemini grounded in advertiser's website |
| Lead Qualification | Manual (by sales team after submission) | Automated (via conversational intent signals) |
| Available Hours | Landing page 24/7, sales follow-up during business hours | Conversation 24/7, pre-qualified handoff to sales |
| Data Captured | Name, email, phone | Same plus questions asked, concerns raised, buying signals |
The advertiser does not build the agent from scratch. Google's system crawls the advertiser's website, ingests the content, and uses it to ground Gemini's responses. The advertiser's optimization task is not prompt engineering or agent design, but ensuring their website content is comprehensive, accurate, and structured in a way Gemini can parse and retrieve.
Why This Matters for B2B Lead Generation
Business Agent for Leads shifts lead generation from a capture mechanism to a qualification mechanism. Traditional lead forms optimize for volume: the fewer fields, the higher the conversion rate, but the lower the qualification signal. Advertisers trade form length for submission rate, then rely on sales teams to qualify leads manually after the fact.
Business Agent inverts this. The conversation itself is the qualification. A prospect who asks five questions about implementation timelines, integration requirements, and pricing tiers has demonstrated far more intent than a prospect who filled out three fields. The sales team receives fewer leads, but each lead arrives with a transcript of what mattered to the prospect.
This also enables a 24/7 sales presence without staffing cost. A prospect researching universities at 11 PM on a Saturday can ask questions and receive answers immediately, rather than submitting a form and waiting until Monday for a response. The lag between ad click and sales contact, which traditionally causes drop-off, is eliminated.
The format raises the intent threshold. Only prospects who engage meaningfully submit. Tire-kickers who would have filled out a form to download a PDF but had no buying intent are filtered out at the conversation stage, because the agent does not pre-fill and submit a form unless the dialogue signals genuine interest.
Current Availability and Eligibility
Business Agent for Leads is available in open beta for US advertisers only as of May 2026. Google tested the format first in three verticals: automotive, education, and real estate. The company has indicated that B2B SaaS, services, and manufacturing categories will follow as the beta expands, but no timeline for international rollout has been announced.
Eligibility is limited to AI Max for Search or Performance Max campaigns with text customization. Standard Search campaigns, Display campaigns, and Video campaigns do not qualify. This means the cost of admission is running Google's AI-first campaign types. Advertisers who prefer manual bidding and traditional Search campaigns cannot access the format.
Google has not publicly disclosed pricing for Business Agent for Leads itself. The format operates within the bidding and budget models of AI Max and Performance Max, but there is no published cost-per-lead benchmark or premium fee structure specific to the conversational agent feature.
The Broader Context: Google's Agentic Advertising Push
Business Agent for Leads is not a standalone product. It is part of a broader shift toward what Google calls the "AI era of Search," where advertising becomes conversational, contextual, and agent-mediated.
Conversational Discovery ads answer specific questions inside AI Mode. A user searching for ways to make their home smell like a spa might see a tailored ad highlighting relevant product features, generated by Gemini in response to the query. The ad is not a pre-written creative unit; it is a real-time response shaped by the user's intent.
Highlighted Answers surface ads as recommendations when AI Mode generates a list. If a user asks for the best language apps for an upcoming trip, relevant, high-quality ads can appear on that list as Highlighted Answers, positioned as recommendations rather than interruptions.
AI-powered Shopping ads, expected to roll out later in 2026, will bring similar conversational mechanics to product discovery. The details have not been fully disclosed, but the pattern is consistent: ads that respond to context rather than simply occupying inventory.
The Brand Safety Question: What Happens When the Agent Gets It Wrong?
Business Agent for Leads hands front-line sales conversations to an LLM. This raises a question Google's announcement materials do not address: what happens when Gemini hallucinates a feature, quotes incorrect pricing, or promises a delivery timeline the advertiser cannot meet?
Traditional lead forms carry no such risk. A prospect fills out fields, submits, and the sales team follows up with accurate information. Business Agent for Leads inserts an autonomous agent between the prospect and the sales team, and that agent is making factual claims about the advertiser's product or service in real time, grounded in website content it crawled and interpreted.
If the agent tells a prospect a software integrates with Salesforce when it does not, or quotes a price that is no longer current, or states a feature is available when it is still in development, the advertiser owns that misrepresentation. The prospect acted on information the advertiser's ad provided. Google has not publicly disclosed liability frameworks, hallucination rates, or advertiser controls for correcting or overriding agent responses.
The format works only if the advertiser's website content is accurate, comprehensive, and unambiguous. Vague marketing copy, outdated pricing pages, or incomplete product documentation will cause the agent to fail or, worse, to confidently state incorrect information. Advertisers running Business Agent for Leads are effectively trusting Gemini to represent their brand accurately based on a crawl of their site. That is a higher-stakes dependency than a static lead form.
What This Signals About Lead Gen's Future
Business Agent for Leads reveals where lead generation is headed. Static forms are being replaced by conversational qualification. The landing page, long the centerpiece of lead-gen optimization, becomes optional. During the initial conversation, the prospect never visits it. The agent answers questions using content drawn from the advertiser's website, but the prospect does not navigate to the site until after the conversation, if at all.
This changes the optimization surface. Most lead-gen advice focuses on landing page design: headline clarity, form field reduction, trust signals, call-to-action placement. Business Agent for Leads skips the landing page entirely during qualification. The new optimization task is ensuring the website content Gemini uses to answer questions is comprehensive, accurate, and accessible. If the agent cannot answer a prospect's question because the website lacks that information, the conversation stalls.
The lead quality metric shifts from "form completed" to "intent demonstrated through dialogue." A prospect who submits a form has signaled interest. A prospect who asks five questions, receives answers, and then provides contact information has signaled intent backed by understanding. Sales teams will prioritize the latter.
First-party data capture expands beyond contact information to include conversational signals. What did the prospect ask about? Which objections did they raise? Which features mattered most? Which details did they ask to clarify? This data informs not just the sales follow-up but product positioning, messaging refinement, and content strategy.
The format also creates a participation barrier. Advertisers who cannot or will not run AI Max for Search or Performance Max are locked out. Google is using access to new ad formats as leverage to push advertisers toward its AI-first campaign types. Advertisers who prefer manual control or who distrust automated bidding lose access to the conversational lead-gen format entirely.
Finally, the format assumes the advertiser's website content is comprehensive and machine-readable. If the site is thin, outdated, or structured in a way Gemini cannot parse, the agent will fail to answer questions, and the format will underperform. This rewards advertisers who have invested in content depth and penalizes those who rely on minimal landing pages optimized purely for conversion.
How Marketers Should Prepare
Business Agent for Leads is in open beta, limited to US advertisers in select verticals, with no confirmed timeline for general availability or international rollout. But the direction is clear: conversational qualification is replacing static forms, and the optimization surface is shifting from landing page design to website content depth.
Marketers who want to prepare should focus on three areas:
Audit website content for completeness and accuracy. The agent's answers are only as good as the content it can retrieve. Pricing pages, product specifications, integration details, implementation timelines, support terms, and edge-case FAQs must be published, current, and unambiguous. Vague marketing copy or incomplete documentation will cause the agent to fail or hallucinate.
Prioritize structured, machine-readable content. While Google has not disclosed whether Business Agent for Leads uses schema markup or specific HTML structure to retrieve content, the general pattern across Gemini-powered features favors well-structured, semantically clear HTML over unstructured PDFs or image-based content. Migrating key information from PDFs into HTML, building out FAQ pages with explicit question-and-answer pairs, and ensuring product data is tabular and machine-parsable will improve retrieval quality.
Test AI-first campaign types now. Access to Business Agent for Leads requires running AI Max for Search or Performance Max. Advertisers who have never run these campaign types, or who distrust automated bidding, will face a learning curve. Testing these formats now, even for a small budget, builds familiarity with how they work and whether they deliver acceptable performance before the conversational format expands.
