Imagine a homebuyer in Plano. They want a four-bedroom under $650K, near a top-rated elementary school, with a pool, in a specific ZIP code. In the old world, they searched for Zillow every morning for six months.
In the new world, they spin up an information agent at Google AI Mode, tell it their criteria once, and walk away. The agent monitors Zillow, Redfin, neighborhood blogs, local agent websites, MLS feeds, school district announcements, and city zoning updates around the clock. When something matches, the user gets a synthesized notification with sources cited.
That homebuyer might search Google once a quarter from now on. Maybe less.
Now imagine being the realtor, the brokerage, or the local blog that agent pulls from when it sends those notifications. Or imagine being the realtor it never pulls from. That’s the new game, and it starts rolling out this summer.
What Google’s information agents actually do
At I/O 2026, Google introduced information agents as part of the biggest Search overhaul in 25 years. The product is straightforward to describe and disruptive to think about.
A user gives the agent a topic, a question, or a set of criteria. The agent scans the web continuously — blogs, news sites, social posts, Google’s real-time finance and shopping data, sports feeds, everything. When the conditions are met, or when something changes, the agent sends a synthesized update with links to the sources it used.
The keynote framed this with sneaker drops and apartment hunting. The real use cases will be bigger.
A personal injury attorney monitoring new case law in their state. A dentist tracking insurance reimbursement changes. A home services franchise watching pricing shifts at competitors. A B2B buyer monitoring SaaS feature releases. A policy team tracking regulatory filings. A patient watching for clinical trial enrollment. The pattern repeats across every vertical we work with at 12AM.
Information agents launch first for Google AI Pro and Ultra subscribers this summer, but the architecture they introduce affects everyone. Because the moment users get used to setting agents instead of searching, the content economy shifts under them.
See exactly where your profile stands right now.
Our GBP audit shows your current rank position across your market, how your profile completeness scores against competitors, and the specific gaps holding you back from the Map Pack.
The shift from search to subscription
Here’s the mental model that helps.
For 25 years, search has been a transactional behavior. Type a query, get results, click, leave. Repeat tomorrow if you have the same question. Every search starts from zero.
Information agents replace that with something closer to a subscription. The user expresses an interest once, and the agent maintains an ongoing relationship with the topic on their behalf. The agent does the searching, the filtering, and the synthesis. The user just consumes updates.
This is closer to how RSS worked in 2005, except the subscription isn’t to a specific feed. It’s to a topic, with the AI choosing which sources to pull from at any given moment.
If your business is one of those sources, you compound. The agent keeps returning. Your brand gets reinforced in the user’s mind over weeks and months. When they’re ready to act, you’re top of mind because you’ve been the source their agent quoted six times in the last quarter.
If you’re not one of those sources, you don’t get a second chance. The agent doesn’t reconsider you next week. It already picked who it trusts.
The three signals agents prioritize
We’ve been studying citation patterns in AI Mode, Gemini, ChatGPT, and Perplexity for a year. The patterns predict what information agents will weight, because they’re built on the same underlying models. Three signals matter most.
Recency
Agents are biased toward fresh information. A page updated in the last 30 days beats a page updated in the last year, even when the underlying content is similar. This is a fundamental break from old-school SEO, where evergreen content could sit untouched indefinitely.
For agent visibility, the content that wins is content that updates. Pricing pages updated as prices change. FAQ pages updated as questions change. “What’s new” pages, changelogs, news feeds, last-modified timestamps that move. Static brochure sites lose. Living sites win.
Structure
Agents can only pull what they can parse. A page with clean headers, schema markup, structured data, FAQ blocks, and direct answers in the first sentence after a question gets pulled. A page with five paragraphs of context before the actual answer gets skipped.
This is where most websites fail. They’re written for humans who’ll scroll. They’re not written for agents that scan.
Authority
Agents weight entity strength heavily. Sites with strong Knowledge Graph presence, consistent NAP, real authors with verifiable credentials, third-party citations, and clean reputational signals get picked. Anonymous sites with no entity layer don’t.
Authority compounds. Once an agent has cited you twice, you become a more trusted source for future queries on adjacent topics. The first citation is the hardest.
The content types that get disproportionate agent attention
If you’re rebuilding content strategy for the agent era, prioritize these.
Living comparison pages. “X vs Y” pages that get genuinely updated as products and pricing change. Agents monitoring purchase decisions pull from these constantly.
Pricing transparency pages. Real, current pricing. Not “contact us for a quote.” Service businesses that publish honest pricing ranges get cited disproportionately. The ones that hide it get skipped.
Status pages and changelogs. SaaS companies have figured this out. Service businesses haven’t. A “what’s new this month at our practice” page, updated genuinely, becomes an agent magnet for users monitoring you and your category.
News and announcement hubs. Press releases hosted on your own domain, with proper Article schema, dated correctly, on a topic-clustered news section of your site. These trigger agents watching for industry changes.
Original data and reports. Anything with real numbers — survey results, internal benchmarks, performance data, market analysis. Agents synthesizing answers love original sources because they reduce the chance of pulling stale or paraphrased information.
FAQ pages with frequent updates. Not the FAQ page you wrote in 2019 and never touched. A genuinely maintained question bank with FAQPage schema, refreshed monthly, with new questions added as they come in from real customers.
What this looks like in our four verticals
Legal practices. A monthly “case results” page with anonymized settlement summaries and dates. A “Texas personal injury law updates” page tracking statutory changes. Both with proper schema, both updated regularly.
Dental practices. A “new patient FAQ” that gets refreshed as questions evolve. A page tracking insurance and PPO changes in your area. A monthly “what’s new at our practice” section covering technology, providers, and procedures.
This is the work we do for you. Every week, without exception.
Managing GBP at this level takes 6–8 hours a week when done right. Nova handles the entire system — posts, photos, reviews, Q&A, citations, heatmap tracking — so you can focus on running your business.
Home services. A “current pricing guide” updated quarterly with honest ranges. A “service area updates” page tracking new neighborhoods. Seasonal advisory content (storm preparation, HVAC pre-season checklists) refreshed annually.
Franchise systems. A “newly opened locations” feed with dates and addresses. A “current franchise opportunities” page with real-time availability. Press release archive with structured news schema.
None of this is complicated. Most of it is content the business should already be producing. The shift is treating it as agent infrastructure, not as marginal blog content.
The agent-readiness audit
Run your site against five questions.
- Which pages on your site have a dateModified timestamp from within the last 90 days? If the answer is “none” or “I don’t know,” your site reads as stale to AI systems.
- Where on your site can a user (or an agent) find current pricing, current service offerings, and current availability? If those answers live in scattered places, agents can’t pull them cleanly.
- Is there a “what’s new” or “latest updates” section anywhere on your site? If not, you’re not signaling to agents that you’re an active source.
- How often is your FAQ page updated? Real answer.
- If an agent were monitoring your industry today, would it have any reason to come back to your site weekly? Be honest.
Most clients fail three or more of these on first audit. That’s the starting point.
How 12AM is building agent-ready content for clients
We’re treating information agents as the primary distribution channel for organic content going forward. That means content calendars built around updateable, citeable, structured assets — not one-time blog posts that go stale the day they publish.
Our approach: identify the 15-25 living pages every client should maintain (pricing, FAQ, comparison, news, status, original data), build them with proper schema and freshness signals, and put them on quarterly maintenance cycles so they keep triggering agents over time.
It’s a different content strategy than what most agencies are still selling. It’s also the one that compounds inside the new Google.
