Google Labs: The emerging content stack

I watched a recent interview between Vaibhav Sisinty, founder of GrowthSchool, and Josh Woodward, VP, Google Labs & Google Gemini. What makes it worth watching is that Woodward walks through a set of public Google AI products and experiments that, taken together, reveal a much bigger shift in how Google wants creative work to happen.

One interview. Seven demos. One much bigger signal.

On the surface, this looks like another executive interview plus product showcase. Underneath, it is a useful snapshot of Google’s current AI surface across content, design, research, image editing, music, immersive world-building, and communication. Google Labs is the home for AI experiments at Google, and the interview makes that portfolio feel less like scattered demos and more like an emerging system.

The setup is simple. One conversation shows how a marketer can move from source material to interface concept to visual asset to soundtrack to presentation layer without switching mental models every five minutes. That is why the interview matters more than the usual AI highlight reel.

Google is no longer just shipping tools. It is sketching a marketing workflow.

A marketing workflow is the connected chain of jobs from understanding a brief to shipping an asset, interface, or experience.

Google’s current AI surface now covers adjacent stages of work that used to require a mess of separate tools. Stitch handles UI design and front-end generation for apps and websites. NotebookLM handles source-grounded understanding. Pomelli handles on-brand marketing content. Nano Banana 2 handles image generation and editing. Lyria 3 handles music creation inside Gemini. Beam extends the stack into communication.

In practical terms, this means more of the work can happen inside one Google-shaped environment instead of bouncing across a pile of disconnected tools.

My view is that Google is not showing isolated AI tricks here. It is sketching the outline of a marketer-friendly workflow it wants to own. The real question is not whether every tool is perfect yet. It is whether Google is making enough of the workflow usable in one place that marketers start changing their habits.

The tools that make the pattern easy to see

Pomelli

Pomelli is the most directly marketer-facing tool in the set. It is built to help businesses generate on-brand content faster. Easy use case: give it your site and product context, then generate campaign-ready visuals and messaging variations for social, ecommerce, or CRM. I unpacked one part of that story in my earlier Pomelli Photoshoot deep dive.

Stitch

Stitch is Google’s answer to fast interface ideation. It turns prompts into UI concepts and front-end output for mobile apps and websites. Easy use case: turn a campaign landing-page idea or app flow into a first working interface before design and dev teams invest heavier production time.

NotebookLM

NotebookLM stands out because it starts from your own source material. It helps turn messy research into usable understanding. Easy use case: upload research docs, interview notes, or previous campaigns and use it to build a grounded strategy summary, FAQ, or narrative draft.

Project Genie

Project Genie is the experimental outlier, but it matters because it points to where interactive creation is heading. It lets users explore generated worlds in real time from simple prompts. Easy use case: prototype a branded world, retail concept, or immersive experience before committing to a more expensive 3D or gaming build.

Nano Banana 2

Nano Banana 2 is Google’s latest image-generation and editing push inside Gemini. It is built for faster visual creation, editing, and iteration. Easy use case: create localized campaign visuals, packaging mockups, or quick ad variants from one approved base asset without opening a traditional creative suite first.

Lyria 3 in Gemini

Lyria 3 brings music creation into Gemini. It lets users generate short custom tracks from prompts and creative inputs. Easy use case: create a first-pass soundtrack or mood bed for a product reel, internal concept film, or social clip before moving into full production.

Google Beam

Google Beam, formerly Project Starline, is the communication layer in this broader picture. It turns standard video streams into a more life-sized and spatial experience. Easy use case: use it for high-stakes remote collaboration, premium client conversations, or executive workshops where trust and presence matter more than standard video calls can deliver.

Why this lands faster than most AI demos

Most AI demos still fail the practical test. They show capability without showing where that capability fits into real work. This one lands because the tools map onto jobs people already understand. Research. Design. Asset creation. Editing. Sound. Presentation. Collaboration.

That is what makes the portfolio more memorable than a long list of model upgrades. People do not buy into AI because a benchmark moved. They buy in when they can picture a job getting easier, faster, or more creatively open.

What Google is really trying to own

Google’s business intent looks bigger than feature adoption. It is trying to make more of the marketer’s daily workflow feel native to its own ecosystem, from idea formation to content generation to communication. That is a stronger strategic position than winning a one-off feature comparison.

This is also why labs.google matters in the story. It is not just a gallery of experiments. It is the clearest public window into which adjacent jobs Google thinks belong together next.

What marketers should take from this now

Do not watch this interview as another AI tool roundup. Watch it as a preview of how Google wants more of the marketer workflow to happen inside one ecosystem.

Extractable takeaway: The strategic signal here is not one impressive Google AI demo. It is that Google is assembling enough connected creative building blocks that marketers can start reducing tool sprawl and shortening the path from brief to output.

The practical move is to start small and test the clearest sequence. NotebookLM for synthesis. Stitch for interface concepts. Pomelli or Nano Banana 2 for visual production. That is already enough to show whether your current bottleneck is research, creative iteration, or production speed.


A few fast answers before you act

Which Google tools in this interview matter most for marketers right now?

NotebookLM, Stitch, Pomelli, Nano Banana 2, and Lyria 3 are the most directly useful because they map to research, interface concepts, asset creation, editing, and soundtrack generation.

Why does this interview matter more than a normal product launch video?

Because it shows multiple Google AI products side by side, which makes the workflow pattern easier to spot than a single product announcement.

Is Google Labs just a showcase site?

No. It is Google’s public home for AI experiments, which makes it the best place to track how Google is connecting adjacent creative and knowledge tasks.

What is the clearest first test for a marketing team?

Use NotebookLM to digest source material, Stitch to mock the experience, and Pomelli or Nano Banana 2 to produce first-pass campaign assets.

What is the strategic takeaway for leaders?

Evaluate these tools as a workflow play, not as isolated demos, because the compounding value comes from reducing friction between connected jobs.

NotCo: AI-Powered Fragrance With Purpose

Back in 2014, Oscar Mayer showed how powerful scent becomes when it stops behaving like a message and starts behaving like a mechanic. Its bacon alarm let people wake up to the sound of sizzling bacon on the stove, while the brand inserted itself into a daily habit instead of a one-off impression.

Fast forward to 2026, and NotCo is pushing scent from playful activation into AI-enabled product development. With Giuseppe AI and its fragrance formulation work with Cramer, a Latin American multinational in flavors and fragrances, NotCo is showing how a sensory cue can become a personalized product proposition. Giuseppe is positioned as an end-to-end product development platform, meaning it helps move from idea to formulation to scalable output within one workflow.

How Aroma Best Friend makes Giuseppe easy to understand

Aroma Best Friend does not try to explain AI through dashboards, technical architecture, or speed claims. It explains the platform through a very human tension point: a dog struggling when its owner leaves home. The story is simple, emotional, and commercially useful at the same time.

The mechanism is easy to retell. The campaign presents a personalized fragrance generated from the owner’s scent profile so a dog is left with an olfactory stand-in for presence. An olfactory profile is the identifiable mix of volatile compounds associated with a person’s scent signature.

In consumer goods, this is the kind of AI story that travels fastest because it links formulation capability to a sensory outcome people can instantly understand.

The film frames the idea around making your dog happier, which keeps the promise focused on an outcome instead of a technology demo.

Why this lands harder than most AI demos

Most AI campaigns still make the same mistake. They tell you the model is powerful and then expect the audience to infer the commercial value. Aroma Best Friend works better because the technology claim is attached to a felt problem and a tangible output, which makes the platform easier to understand and easier to remember.

Extractable takeaway: AI becomes more persuasive when it is shown solving a problem people can emotionally grasp, not when it is described as a capability stack. The sharper the human tension and the clearer the output, the stronger the commercial story.

Scent is not decorative here. It is the proof. That turns Giuseppe from a backstage R&D engine into the source of a new kind of product experience. NotCo is not just advertising AI. It is advertising the kinds of product experiences AI can now help create.

The business play behind the emotion

The real question is whether an AI platform can turn an invisible R&D capability into a story that brand teams, partners, and future buyers instantly understand.

The official waitlist for the product makes clear that joining does not guarantee access to or availability of the product. That suggests this is as much about validating demand and capturing interest as it is about launching a ready-to-scale offer.

That is the smarter move. Aroma Best Friend works as a campaign, a proof-of-capability demo, and a demand signal test at the same time. Instead of saying that Giuseppe enables personalization and creativity, NotCo dramatizes a specific version of personalization that people can picture, repeat, and remember.

What FMCG and CPG teams should borrow now

  • Turn capability into consequence. Do not market the model first. Market the human outcome the model makes possible.
  • Use one emotionally legible use case to explain a broader platform. Aroma Best Friend is about dogs on the surface, but the deeper message is that Giuseppe can work where formulation and personalization matter.
  • Make the demo do double duty. The strongest AI campaigns are not just communications assets. They also test demand, capture leads, and reposition the company.
  • Choose outputs people can feel, not just read about. Text is easy. Fragrance is harder. That is exactly why this idea carries more weight.
  • Prove customization through specificity. Personalized fragrance is stronger than generic AI-powered personalization because it gives the claim an object, a use case, and a memory.

A few fast answers before you act

What is Aroma Best Friend really marketing?

Aroma Best Friend markets a personalized scent concept for pet separation anxiety on the surface, but at a deeper level it markets Giuseppe AI as a product-development engine that can move into formulation-led use cases.

Why does this explain Giuseppe better than a typical AI demo?

It explains Giuseppe better because it connects the technology to a human problem and a sensory output. That makes the platform easier to understand than abstract claims about intelligence, speed, or creativity.

Is Aroma Best Friend already a scaled product launch?

Not yet in any proven commercial sense. The waitlist language makes clear that joining does not guarantee access to or availability of the product, so the initiative still functions as a signal test as much as a launch story.

Why is scent such a strong choice for this idea?

Scent carries memory, comfort, and presence more directly than most brand cues. That gives the campaign emotional force and turns formulation technology into something people can instantly imagine in use.

What should marketers and innovation teams steal from this?

They should steal the structure. Start with a real human tension, let the technology solve it in a tangible way, and make the output specific enough that people can retell the story in one sentence.

Super Bowl 2026 Ads: By Ad Recall

It’s been two weeks since the Super Bowl, but the most important data from advertising’s biggest night lands now, after the noise has died and the industry has moved on to arguing about something else. Ipsos recall data shows that long-running campaigns outperform bespoke event ads by embarrassing margins.

The only Super Bowl signal that survives Monday

Every year, the game becomes a weeklong festival of hot takes, rankings, and creative commentary. The game itself produces a clear winner, but in the industry we speak too generally about Super Bowl advertising as if it’s all the same. It isn’t.

That’s the problem. We talk about “Super Bowl advertising” as a category, when the night produces winners and losers in advertising too.

How Ipsos turns hype into a memory test

Ipsos tracked spontaneous brand recall among Super Bowl viewers, the simplest and most demanding test in advertising. Viewers were asked which brands they remembered seeing advertised during the game, with no prompts. Ipsos measured it the next morning, and again a week later.

Spontaneous recall is unaided naming. If people cannot name you without a list in front of them, you were entertainment, not advertising.

In global FMCG and retail portfolios, tentpole moments are recurring, so the only scalable advantage is a set of distinctive brand cues that work across every channel.

The real question is whether your Super Bowl spot is building durable brand memory or renting a one-night reaction.

The winners did not act like it was a one-night event

Budweiser

Budweiser dominated the night on recall. Its “American Icons” spot, a foal and a newly hatched bald eagle growing up together over the years to Lynyrd Skynyrd’s “Free Bird”, achieved spontaneous recall of nearly twenty million viewers the morning after the game.

A week later, that number had climbed to twenty three million. Perhaps a signature of a campaign that made it into actual memory rather than mere social feeds.

Pepsi

Pepsi came second, 12 million viewers still able to recall the brand the next day. Their polar bear blind taste test, Coca-Cola’s own mascot choosing Pepsi over Coke, landed because it was built on decades of competitive positioning and the oldest tactic from the cola wars: the challenge.

Dunkin

Dunkin finished third at 11 million. Ben Affleck’s star spangled sitcom parody is exactly what good advertising should be: emotionally engaging, distinctively coded, impossible to misattribute.

These three brands make a compelling case that Super Bowl advertising can work. Spend well, follow your strategy, and put your name into millions of minds and leave it there for a week. Factor in a hundred million viewers, the additional coverage, social amplification, and the required $8 million investment looks seriously worth it, particularly in the fragmented, chaotic media landscape we now inhabit.

The expensive part is not the media. It is the forgettability

The uncomfortable half of the Ipsos results is how many brands barely cleared the minimum bar. More than half the brands in the Ipsos data gained less than a percentage point of recall the morning after their ad ran. Each spent what most companies deploy as an entire year’s marketing budget. Each one has very little to show for it.

Ring

Ring managed 26th place with less than one million viewers recalling the brand the next day, roughly a twentieth of Budweiser’s number. Recall picked up later, likely driven by the outcry around the AI narrative in the spot.

Michelob Ultra

Michelob Ultra came 44th out of 45 brands after running a glossy, star studded spot featuring Kurt Russell, Chloe Kim, and T.J. Oshie. It cost a packet to produce, north of $8 million to air, and was instantly forgotten by almost every viewer of the big game. While recall improved a week later, that was most likely “ghost recall” from other spend rather than the Super Bowl moment itself. By “ghost recall,” I mean recall created by other media exposures that gets wrongly attributed to the Super Bowl airing.

It’s perhaps unfair to single out two brands when almost two thirds of those advertising during the Super Bowl failed so miserably to reach even the lowest bar in the persuasion hierarchy.

Why “familiar” beats “fresh” in 30 seconds

So what separates the winners from the losers? It’s mostly a story of consistency. Ring and Michelob Ultra made special Super Bowl ads. Budweiser and Pepsi didn’t. They extended long running brand codes into the Super Bowl opportunity. It’s not a small distinction.

Extractable takeaway: If your Super Bowl idea does not extend an existing set of brand cues, assume you are buying applause, not memory.

This was the 48th time the Budweiser Clydesdales appeared during the game. Clydesdales are large draft horses, and Budweiser has used them for decades as a signature brand cue in its advertising. Forty-eight years of the same visual assets and the same emotional territory. Think of it the other way: decades of ignoring hot agencies and ambitious new CMOs wanting to “put their stamp on things.” Either way, the sight of those horses trotting across a field now makes 20 million people think of one beer and one beer only.

“Distinctive assets” are repeatable cues, characters, music, visual codes, and phrases that people reliably link to a specific brand. When those cues repeat, viewers identify the brand faster and more accurately, which increases the chance the story is stored as brand memory rather than background entertainment.

Business intent. Buy memory, not applause

Most marketers know patience wins. But very few act on it, because patience is not rewarded in quarterly business cycles and it certainly won’t win many industry awards.

Our industry is structurally biased toward newness. Marketers want to make new ads, and agencies, who get paid to create new work and nothing to run the old, aren’t incentivized to argue with them.

Some brands use the biggest advertising night of the year to launch something bespoke, something special, something that will live nowhere after the post game debate ends.

Budweiser used it to add one more chapter to something it started building long before today’s marketing teams rotated in. The Clydesdales are not a campaign. They are compound creativity, and compound creativity is what memory looks like.

Steal this from the recall winners

  • Keep the Super Bowl brief brutally narrow. Your first job is correct attribution, then entertainment.
  • Write an “asset continuity brief” before the creative brief. List the 3 to 5 cues you will not change.
  • If you make a one-off Super Bowl ad, brand it hard. New characters plus subtle branding is the fastest route to being forgotten.
  • Measure decay, not just peak. Next day recall is the entry ticket. Day 7 tells you whether you made memory.
  • Build for reuse. If the idea cannot live beyond one night, it is a very expensive dead end.

A few fast answers before you act

What is “spontaneous brand recall”?

Spontaneous brand recall is an unaided memory test. People are asked which brands they remember seeing advertised, without being shown options.

Why do long running campaigns usually win on recall?

Because repeated cues let viewers identify the brand quickly and correctly, which makes it more likely the story and brand get stored together in memory.

Does this mean you should never make a special Super Bowl ad?

No. Make the story special. Keep the brand cues consistent.

What is the fastest pre flight test before you approve the spot?

Ask neutral people: “Who is this for?” If they cannot name the brand quickly, the work is at risk.

What should you track besides recall?

Correct brand attribution, brand lift, search lift, and any downstream sales proxy you trust. Recall is the first gate, not the finish line.