March was a busy month for AI in advertising. AI shopping agents moved from conference demos into real consumer behavior, Google shipped new image generation tools and gave a candid interview about the future of ads in Gemini, and the data on AI Overviews’ effect on search clicks kept piling up. Here are the five developments online advertisers should be paying attention to.
1. AI Shopping Agents Are Here, and They’re Already Changing How People Buy
Sources: TechCrunch | March 16, 2026; NVIDIA | March 16, 2026
Agentic shopping moved from theory to practice this month. Shopify president Harley Finkelstein told an audience at the Upfront Summit that the company is going all in on AI shopping agents, calling them the future “personal shoppers” that will serve as a new front door for e-commerce. Shopify is building protocols so agents can understand merchant product data and act on behalf of consumers. Meanwhile, NVIDIA announced NemoClaw at GTC 2026, an enterprise-grade runtime that adds security and privacy guardrails to open-source AI agents like OpenClaw. The platform, which has racked up over 135,000 GitHub stars, lets users automate tasks through messaging apps like Telegram, WhatsApp, and Slack. At the retail level, Target is collaborating with OpenAI, and Walmart has partnered with Google’s Gemini on agentic commerce initiatives.
| JumpFly Takeaway for Marketers This is something I’ve been testing personally. Recently, I used an OpenClaw-style agent through Telegram while I was driving. I sent a message asking it to find a specific board game at my local Target. Within minutes, the agent browsed available inventory at that store, identified the exact aisle location, pulled up pricing and customer ratings, and even flagged a current sale. I never opened Target’s app, never visited Target.com, and never saw a single ad. The entire shopping journey happened inside a messaging app. Agentic shopping is only going to evolve from here. The practical move for advertisers is to focus on what these agents actually consume: structured data. A well-organized product feed, accurate Merchant Center data, and an easy-to-navigate site are what put you in front of an AI agent’s recommendation. As more consumers let agents do the browsing for them, the brands with clean data will be the ones that get surfaced. |
2. Google Rolls Out Nano Banana Pro, Its Most Advanced AI Image Generator, Directly in Google Ads
Source: Google Business | March 2026
Google has made Nano Banana Pro available to all active Google Ads users through Asset Studio. The model lets advertisers generate photorealistic product images using conversational editing, meaning you describe the changes you want in plain language and the AI delivers them. Google says Nano Banana Pro can render fine details like text on product labels and packaging, maintain product fidelity from original assets, and compose scenes featuring up to five products in a single lifestyle image. The tool also supports creating seasonal ad variations from existing top-performing creatives without rebuilding from scratch. Early adopters report using it to rapidly test creative hypotheses that would previously have required a full design cycle.
| JumpFly Takeaway for Marketers Having Nano Banana Pro directly in Google Ads is a potential game-changer (this is my catchphrase). With the Pro version of Nano Banana, you can iterate on creative quickly, stay within one platform, and there’s no extra cost to use it. We run a lot of Performance Max and Demand Gen campaigns, which are creative-heavy by nature, so anything that speeds up asset production without sacrificing quality gets our attention. Being able to describe the image you want in plain language and getting a usable result is a step beyond what previous AI image tools have offered. We’d still recommend reviewing every generated image before it goes live, but blend this into what you’re already doing, and let it shorten your testing cycles. |
3. AI Overviews Are Quietly Eating Search Clicks
Source: Semrush | Updated March 2026
The data on AI Overviews is getting harder to ignore. When AI Overviews appear on a search results page, click-through rates on the listings below drop by roughly 42%, according to Semrush research. AI Overviews now show up for about 13% of all Google queries, and ads appear alongside 25.5% of those AI Overview results, up from just 5.17% earlier in 2025, a 394% increase. At the same time, Google VP of Ads Dan Taylor confirmed that ads within AI Overviews are now monetizing at the same rate as traditional search ads, and Q4 2025 search revenue climbed 17% year over year. The zero-click trend is accelerating: users are getting answers directly from the AI-generated summary and never scrolling down. For advertisers and SEO-driven brands alike, the math on organic traffic is changing.
| JumpFly Takeaway for Marketers I’ve personally seen numerous accounts drop in organic search sessions year over year. It’s hard to look at that data and blame poor management when the timing lines up so neatly with AI Overviews rolling out. Users are getting their answers right there in the AI summary and never clicking through. Having the right SEO and AEO (AI Engine Optimization) strategy is no longer optional. This is not going away. If your content is structured to answer questions directly, with clean product data and authoritative information, you have a much better shot at being the source an AI Overview pulls from. If your SEO strategy has been focused purely on traditional blue-link rankings, it’s time to broaden your thinking to include how your content appears inside AI-generated answers, not just below them. |
4. Google Is Not Ruling Out Ads in Gemini
Source: WIRED | March 2026
In an interview with WIRED, Google SVP Nick Fox said the company is “not ruling out” ads in the Gemini app. Fox noted that learnings from the company’s ongoing ad experiments in AI Mode would “likely carry over” to Gemini down the road. Google is currently testing ads in AI Mode, its search experience powered by Gemini, but the Gemini standalone app remains ad-free for now. Fox described a three-part approach to AI and advertising: better measurement (understanding how AI is already driving value), better advertiser tools (AI helping small businesses build campaigns), and ads in new experiences, which he called “the most nascent” of the three. Separately, Google confirmed that users who enable Personal Intelligence, a feature that connects Gmail, Photos, and YouTube to AI Mode, will not see ads in that experience.
| JumpFly Takeaway for Marketers My gut says it’s only a matter of time before ads show up in Gemini. Google has too much incentive and too much infrastructure not to go there eventually. The Personal Intelligence detail is interesting, though. Google is creating two tiers within AI Mode: one with ads, one without. That mirrors what we’ve seen with ad-supported vs. premium tiers on streaming platforms. For now, there’s nothing to change in your campaigns. But we’ll be here waiting to capitalize on marketing within the platform the moment it opens up. |
5. 63% of U.S. Adults Say Ads in AI Search Make Them Trust Results Less
Sources: eMarketer | March 2026; Ipsos | 2026
An Ipsos survey found that 63% of U.S. adults say ads in AI search results make them trust the results less. Fewer than a quarter said ads had the opposite effect. This data sits behind one of the bigger strategic splits in the AI industry right now. Perplexity, the AI search startup valued at $18 billion, has effectively abandoned its advertising program, with its ad chief departing within nine months and the company saying it may never need ad revenue given its subscription growth. Meanwhile, OpenAI is moving in the opposite direction, actively building out an ad infrastructure for ChatGPT. Google is threading the needle with its Personal Intelligence play, offering an ad-free AI experience for users who connect their data. The question the industry is wrestling with: can ads coexist with trust in AI-generated answers?
| JumpFly Takeaway for Marketers I must fall in the other 37%. Here’s the reality: most people can’t afford to pay for every AI platform they use. Ads are what keep these tools free. And personally, I’d rather see ads that are relevant to me than pay another subscription fee. It’s not like the ad is going to be blatantly dropped in the middle of an AI response, because if it were, users would stop using the tool. These companies know that. The ones that get the ad experience right, where it feels helpful instead of intrusive, will keep their users. For advertisers, the takeaway is to keep testing these placements where/when they’re available. ChatGPT Ads are just getting started, and Google is already showing ads in AI Overviews; we just cannot see the attributed data (yet). The platforms that maintain user trust will be the ones worth investing in long-term. |
Looking Ahead
The through-line this month is that the surface area where traditional ads work is shifting. AI agents are finding products through messaging apps. AI Overviews are answering questions before users click, and this number continues to grow. And consumers are telling pollsters they don’t trust ads mixed into AI results. The core principles haven’t changed, though. Clean data, strong product information, and real authority still win. How those signals get delivered to buyers is just evolving faster than it has in a long time.
Follow along each month as we share our insights into the latest AI trends. You can read our previous blogs: 5 Key Trends from January 2026 and 5 Key Trends from February 2026.
If you missed last month’s post, read our article, AI in Online Advertising: 5 Key Trends from January 2026.
Questions about how these trends affect your campaigns? Let’s talk.
