Google’s biggest ads event of the year happened today. Strip out the hype and there are five concrete shifts a small marketing team needs to plan for in the next 90 days.
Today, May 20 2026, was Google Marketing Live, Google’s annual showcase for everything new in Google Ads, YouTube, and Shopping. This year was almost entirely about agentic advertising. Five announcements stood out for small and mid-market marketing teams: agentic commerce (AI assistants shopping on behalf of consumers), Smart Bidding Exploration (automated bid testing), a Gemini-powered reporting dashboard you can prompt in plain English, YouTube shoppable formats with Buy with Google Pay checkout, and journey-aware bidding that finally accounts for mid-funnel touchpoints. Here is what each one actually means, who should care, and what to do about it in the next 90 days, without spending a single extra dollar on tools.
Today, May 20 2026, Google held its annual Marketing Live keynote at 8:45 AM Pacific. The European session runs tomorrow at 11:00 AM BST. [1]
If you have been to one of these in person you know the routine. Big stage, polished demos, four hours of “the AI is going to do it for you” framing. If you watched the livestream, you probably came away with two reactions. One, this is genuinely impressive. Two, what do I actually do with this on Monday?
I have spent the last decade running paid teams and watching every major ad platform event since the launch of Performance Max in 2021. Here is what I actually took away from today, filtered through the lens of someone who has to make this work on a real budget with a small team.
Five things to plan for. Skip the rest of the keynote.
Every Marketing Live keynote sells you the future. The real job is figuring out which two or three pieces of that future change what you do this quarter.
The flagship announcement was agentic commerce. In plain English: Google is building AI agents that shop on behalf of users. Instead of you searching for “best running shoes for flat feet,” an agent in your Google account does the research, compares options, reads reviews, and presents you with two or three filtered choices. Eventually it will check out for you. [2]
Here is what this changes for marketers. If part of your traffic is going to be AI agents instead of humans in the next 12 to 18 months, the things that win clicks are different.
Agents do not respond to:
Agents do respond to:
If your product pages and ads are written like a 2018 lifestyle landing page (one big hero image, three bullet points, lots of vibes), the AI agent is going to skip you in favour of the brand that publishes a complete spec sheet. The companies that already write for AI assistants (we covered the basics of this in our brand citation playbook) are about to pull ahead.
What to do this week: open your three top product pages. Are the specs, materials, sizes, returns policy, and shipping times stated explicitly in text on the page? Not in a PDF. Not in a modal. On the page. If not, fix that first. Everything else is downstream of that.
Smart Bidding Exploration is the new feature that turns Google’s bidding algorithm into a continuous experimenter. Instead of you setting up manual experiments to test “what happens if we bid higher on long-tail queries,” the system runs micro-experiments inside your campaign and reallocates spend toward what worked. [3]
If you have used Performance Max for the last two years, this is the logical next step. The bidding side is becoming fully autonomous.
What I have seen in early access: campaigns running Smart Bidding Exploration find efficient pockets of demand the manual setup misses, especially for accounts that historically underspent on broad match. The downside is the same as Performance Max: less control, less visibility into why a decision was made.
What this means for small teams: the value of your time as the “person who optimises bids” just dropped. The value of your time as the “person who decides the offer, the creative, and the audience strategy” just went up. If you have been spending two hours a week tweaking bid adjustments by device or time of day, that time is better spent on creative and landing pages. Let the machine bid.
Google previewed a Gemini-powered reporting layer inside Google Ads. You can now ask questions like “what changed in our conversion rate last week and why” or “which products underperformed in May compared to April” and get a written answer with the underlying data attached. [4]
This sounds like a small thing. It is not. If you currently spend Friday afternoon or Monday morning pulling numbers from Google Ads into a Slides deck to send to your CMO or your client, that work is about to be automated.
I am not saying you should fire your reporting analyst. I am saying you should redirect their time. The reports are about to write themselves. The work that still needs a human is interpreting the answer, deciding what to do about it, and making the case for a budget shift. That is the work that gets harder to outsource and more valuable to do well.
What to do this week: next time you write a weekly performance report, draft the analysis first, then ask Gemini in Ads to pull the supporting numbers. Notice how much of the work was “explain what happened” versus “stare at a pivot table.” The first part is your job. The second part is going away.
YouTube updated two things in the keynote. First, shoppable formats are now available in connected TV ads (the ads people watch on their actual TV through YouTube). Second, Buy with Google Pay checkout is now integrated directly into those formats, meaning a viewer can complete a purchase without leaving the YouTube experience. [5]
Why this matters more than the average “we made the funnel shorter” announcement: the entire chunk of dropoff between “saw an ad on TV” and “found the product on my phone and bought it” just collapsed for the people Google can identify across devices. For consumer brands with strong YouTube presence, this is a real shift in CAC math.
For B2B and service businesses, it is mostly noise. Do not let the Twitter hype convince you that YouTube is suddenly your channel if it was not before.
What to do this week if you sell physical product: talk to your YouTube account rep (or your agency) about which of your existing video ads can be repurposed into shoppable connected TV formats. The biggest unlock for most brands is not making new creative. It is enabling new placements on what you already have.
Journey-aware bidding is the announcement nobody is talking about that I think will matter most. It is a change to how Google attributes value across the customer journey. Instead of last-click dominating, the bidding system now assigns weight to mid-funnel touchpoints (like a YouTube view that did not result in a click but did precede a search) and bids accordingly. [6]
If you have been running brand awareness or thought leadership content and watching the algorithm undervalue it because nobody clicks immediately, this is the long-overdue fix. The mid-funnel content is finally going to look like the conversion engine it actually is.
It also changes how you structure your campaigns. Top-of-funnel and bottom-of-funnel separation was the right call when last-click ruled. Once journey-aware bidding goes mainstream, blended campaigns that nurture and convert in the same budget start outperforming the rigid separation.
This connects directly to what is actually working in B2B AI content marketing right now, where the smartest teams are already measuring assisted conversions and influenced pipeline alongside the click-attributed numbers.
Look at the five announcements together and a pattern jumps out. Every single one of them automates something a marketer used to do by hand.
If your marketing team’s value sits in any of those five areas (bid tweaking, manual A/B tests, weekly reports, attribution modelling in spreadsheets), the next 12 months are going to be uncomfortable. The work is being abstracted away.
If your team’s value sits in choosing the right offer, writing creative that actually converts, building an editorial point of view, picking the strategic priority that nobody else has spotted, then this is your moment. The automation handles the mechanics. You handle the judgement. The teams who already operate this way are going to look like geniuses in 18 months.
The boring 60% of marketing is being automated. The remaining 40% (judgement, creative, strategic taste) is now the entire job.
Strip out the keynote noise. Here is a 90-day plan a small marketing team can actually execute.
Weeks 1 to 2: audit your top 10 product pages or landing pages for AI-readability. Specs, pricing, availability, returns, all in plain text on the page. Hire someone or block half a day on your own calendar to fix the gaps.
Weeks 3 to 4: turn on Smart Bidding Exploration on one of your existing campaigns (start with a non-critical one). Watch what happens over two weeks. Compare to your previous baseline. Decide if you trust it enough to expand.
Weeks 5 to 6: use the Gemini reporting dashboard for your next two weekly reports instead of pulling numbers manually. Time yourself. Use the time you save to write actual strategic recommendations instead of bullet-pointing performance numbers.
Weeks 7 to 9: if you sell physical product, talk to your YouTube rep about shoppable CTV formats on existing video creative. If you sell B2B services, ignore this step and use the time on the next one.
Weeks 10 to 12: rebuild your campaign structure to align with journey-aware bidding. Most teams have artificially separate “awareness” and “conversion” campaigns. Test merging two of them into a single nurturing campaign and let the algorithm do its job.
That is it. Five real changes in 90 days, mapped to today’s keynote. None of them require buying a new tool. All of them require you to stop doing some of what you used to do and start doing different work. Which is exactly what every Google Marketing Live keynote has been telling us for the last three years, only this year they actually mean it.
Google Marketing Live 2026 was held on Wednesday, May 20 2026, starting at 8:45 AM Pacific Time. The EMEA-focused event runs Thursday, May 21 at 11:00 AM BST. Both are available on demand on googlemarketinglive.com after the live broadcast.
Agentic commerce is Google’s term for AI agents that shop on behalf of consumers. Instead of a person searching, comparing, and clicking, an AI agent (typically Gemini-powered) does the research, filters the options, and presents a shortlist. For marketers, this means a portion of future traffic will be AI agents, not humans, and ad copy and product pages need to be structured for both.
No. Start with one campaign that is not mission-critical, run it for two to four weeks, and compare to your previous baseline. The algorithm needs time to learn, and rushing the rollout across every campaign at once is how teams lose visibility into what is working. Pilot, measure, expand.
No, but it changes what they spend time on. The dashboard removes most of the manual data-pulling and weekly report writing. Your team’s time should shift from reporting numbers to interpreting them and making strategic recommendations. The reporting layer is automation. The thinking is still human.
No. Small advertisers actually benefit more, because they typically lack the budget for complex attribution modelling tools. Journey-aware bidding brings sophisticated cross-touchpoint attribution into the default Google Ads experience for free. Mid-funnel content (YouTube views, display impressions, brand searches) finally gets credit toward conversion without you having to build a custom model.
This article was researched and written by Hina for Future Factors AI. Sources include Google’s official Marketing Live 2026 keynote, ALM Corp’s agency analysis, The Keyword’s advertiser briefing, PPC News Feed’s comprehensive overview, and Precise Impact’s small business take. All statistics are sourced and linked in the citations below.
Weekly tips. Real examples. Practical help for busy professionals.
We care about your data, check out our privacy policy.