Updated August 2026: the access requirements and rollout status in this post have been revised based on our own firsthand AI search testing, which found narrower availability than early coverage suggested.

Honest uncertainty from someone researching this daily. Here's what I'm still watching in AI search and why confident predictions about it should make you suspicious.

There's a version of this post I could write where I wrap the series up with confident predictions about where AI search is headed and what venues should do to prepare. That version would be cleaner and probably more satisfying to read!

It would also be less honest than you deserve.

The AI search landscape is genuinely unsettled right now. I've spent months researching this, preparing for a panel on it at NIVA, and running tests with my own team. There are things I know with confidence; those are what the previous fifteen posts are built on. And there are things I'm still watching, still testing, still uncertain about. I'd rather share those openly than paper over them with authority I don't have.

Here's what I'm tracking.

GOOGLE AI MODE AND THE LONG-TERM ORGANIC TRAFFIC QUESTION

AI Mode is real and it is taking organic clicks from websites, including venue websites. What I don't know yet is how much of that traffic loss is permanent versus how much represents a behavioral shift that venues can adapt to by focusing on citation rather than ranking. I've been watching my clients' Google Search Console data closely since the May 2026 core update completed, and the picture is mixed enough that I'm not ready to declare a direction.

Some query categories are more affected than others; event and venue discovery searches are showing different patterns than informational queries. I'm watching this monthly and will write about what the data shows.

One thing worth naming clearly: AI Mode is now the global default, not an optional tab. As of the May 2026 I/O announcement, Google made AI Mode the primary search experience across desktop and mobile worldwide. There are now three distinct layers in Google search:

1. AI Overview (the generated summary box above your blue links, Google's official term for it),
2. AI Mode conversational experience (which is increasingly the whole search interface) and,
3. Traditional blue links below.

They share the same underlying Gemini 3.5 Flash model (announced May 19, 2026) and the same optimization approach. You don't need a separate strategy for each one; you need to be visible to the system that powers all three.

GOOGLE'S PREFERRED SOURCES FEATURE

Google launched Preferred Sources in August 2025, letting users indicate websites they'd prefer to see in their AI-generated results and Top Stories. Top Stories features globally across all languages, and can also appear in AI Mode and AI Overviews in all languages and locales where those features are available. One eligibility detail worth flagging: Preferred Sources only applies at the domain or subdomain level. Per Google's documentation, a subdirectory like www.example.com/blog isn't eligible on its own, the qualifying unit is the whole domain or a subdomain. I went back to the primary documentation on this one because the early coverage was imprecise, and the actual picture is more specific than most summaries suggest.

Here's what it actually does and doesn't do for venues. There's no content-type or publishing-frequency restriction in Google's documentation. Any qualifying domain or subdomain can be included, regardless of what kind of content it publishes. In practice, badging shows up most often on fresh, news-style content, so a venue whose only web presence is a static events calendar is unlikely to see much benefit from this feature specifically. That's a practical limitation, not a documented rule.

However, the feature isn't limited to news publishers. Any site that publishes fresh content regularly can qualify, including venue sites. A venue that publishes show announcements as actual crawlable pages (not just widget-embedded event listings), artist spotlights, or local scene coverage can appear badged when a fan searches related topics. The mechanism works; the barrier is content production, which for most independent venues is the real constraint.

The honest framing: Preferred Sources is a theoretical opportunity for venues with a content publishing habit and largely irrelevant for venues whose only web presence is a ticket calendar. For the NIVA audience specifically, most operators fall into the second category. I'd file this under "worth knowing, not worth prioritizing" unless your venue is already publishing regularly, in which case claiming the badge costs nothing and may incrementally improve your visibility when fans who follow you search relevant topics.

BING'S QUIET DESKTOP SHARE GAINS

Bing gets treated as an afterthought in most AI search conversations because its overall market share is modest. But on US desktop specifically, Bing is sitting at roughly 13.16% market share as of May 2026 and cross-device US share is about 9.82%, which is not nothing (note that measurement firms diverge widely on this). Microsoft has integrated AI deeply into Bing through Copilot, and Bing's event data pulls from Bandsintown, the same infrastructure we covered here.

For venues whose audiences skew older or corporate, and whose fans are more likely to be on a desktop at work than on a phone, Bing visibility may be worth more than most venue marketers are accounting for. I renewed my Google Ads certification recently and was genuinely frustrated by how single-platform the industry's thinking has become. The exam assumes Google is the only search surface that matters. It isn't.

THE ATTRIBUTION PROBLEM FOR AI-DRIVEN CONVERSIONS

This is the one that keeps me up at night, professionally. If a fan finds your venue through an AI-generated answer, searches your name directly, and buys a ticket through your site, that conversion shows up in your analytics as direct traffic or branded search. AI gets no attribution credit; you have no way of knowing the discovery step happened at all.

I've been thinking about this in the context of a client whose conversion tracking I spent months rebuilding, and the AI attribution gap is genuinely harder to close than the UTM problems we were solving there. I don't have a good answer yet. I'm watching whether any analytics platforms build a solution for this, and in the meantime I'm defaulting to the position that if your AI visibility infrastructure is solid, the downstream conversions will follow, even if you can't see them in the data.

THE AGENTIC BOOKING LAYER AND WHY IT CHANGES THE STAKES

This one is no longer something I'm watching from a distance. It's live.

Since November 2025, Google's AI Mode has been able to search for event tickets on a fan's behalf across Ticketmaster, StubHub, SeatGeek, and Vivid Seats, the four platforms Google has named as partners and surface real-time options with direct booking links. As of our testing in August 2026, the full cross-platform booking experience is still limited to Search Labs participants rather than all US users. It's not a separate app you opt into. It's built into AI Mode, though access is gated while this feature is in Search Labs as an experiment. It requires a personal Google account, a US location, Web & App Activity turned on, and as of our most recent testing, it may depend upon a paid Google AI Pro or Ultra subscription. When it does run, it doesn't require the fan to visit your website at all. They can ask "find me two standing floor tickets for a show at [your venue] this month" and Google's agent goes and finds them.

I'm still watching which venues' events actually surface in that agent's results, and what determines the ranking. The agent draws on platforms like Bandsintown, Ticketmaster, and SeatGeek for its data. If your events aren't in a system that agent can read, they're invisible to it. But the specific weighting and ranking logic isn't public yet, and I'm treating this as a watch item until I've seen enough real-world results to say anything definitive.

What I can say now: the case for having your event data in Bandsintown Pro and your GBP events current just got significantly stronger. Those two data sources are the most accessible inputs into the systems Google's agent is drawing on. Waiting on this one isn't a neutral position.

"I'd rather share what I'm still working through than paper over it with authority I don't have."

I'm also watching whether Google's AI Mode fully displaces traditional search results or finds a more stable equilibrium where both coexist for different query types. My working assumption is the latter, but I've been wrong about Google's pace of change before. Anyone who tells you with certainty how this settles is either more informed than I am or less honest.

What this means practically: the recommendations in this series are grounded in what's working now and in the structural fundamentals that have held across multiple platform shifts. Crawlable content, consistent citations, accurate schema, direct audience channels aren't bets on a specific outcome. They're the foundation that keeps working regardless of which direction AI search evolves.

The next four posts are deep dives, where we go further into the technical and strategic layers for operators who want to build something durable. They're longer, more specific, and more actionable than anything in this series so far. The first deep dive starts with the crawlability problem and doesn't leave until it's solved.