← All Conversations
July 28, 2026

Google AI quietly killed my whole content team

AI content creation Google Gemini automation productivity brand consistency

Summary

Google’s recent AI rollout replaces an entire content team by chaining together Gemini, NotebookLM, Nano Banana, and custom Gems into a single, brand‑aware workflow that turns raw files into finished blog posts, ad creatives, product videos, social posts, and even a full website in minutes rather than weeks. The core insight is that persistent brand memory—not just isolated prompts—is the key to consistent, on‑brand output at scale. By loading a company’s hero products, customer insights, and visual assets into a NotebookLM notebook that lives inside the Gemini app, the system “remembers” the brand once and applies that knowledge to every subsequent generation, eliminating the need to re‑explain style, tone, or product details on each request. The workflow begins with a content brief skill attached to Gemini, which pulls from the notebook to produce a fully formatted blog post complete with internal linking. Switching to image mode (or Canvas mode) triggers Nano Banana to generate feature banners and supporting images directly from the notebook’s product photos, guaranteeing visual consistency. The same notebook can instantly spin up lead‑magnet PDFs or retail line sheets. Nano Banana’s iterative capabilities let users refine images via sketch‑based edits, run automatic A/B tests on backgrounds and moods, and resize creatives for different channels in seconds—all starting from a draft rather than a blank canvas. To avoid repetitive prompting, users create Gems: custom Gemini instances pre‑loaded with the brand notebook, specific tools (e.g., Nano Banana for ad shots), style guides, and default instructions. A Gem for studio shots, another for lifestyle images, and a third for campaign copy can be shared across the team, turning prompt engineering into a reusable standard. The result is a self‑sustaining content engine where a single individual can produce what previously required a dedicated writer, designer, video editor, and social manager. Actionable takeaways: 1) Build a central NotebookLM notebook with all brand assets and insights. 2) Use Gemini with a content‑brief skill to draft long‑form copy and PDFs. 3) Leverage Nano Banana for image creation, editing, A/B testing, and resizing. 4) Encode repeatable tasks into Gems and distribute them to eliminate prompt drift. 5) Adopt this stack to cut production cycles from weeks to minutes, dramatically lowering labor costs while maintaining brand fidelity.

Hey

Two years ago I paid a small team to make content.

One person wrote the blog posts. One designed the ad creatives. One edited the product videos. One ran the social campaigns.

And every single piece took days, sometimes weeks, to move from idea to published.

I remember staring at an invoice once and thinking there has to be a faster way to do all of this.

Last week I sat at Google I/O and watched the faster way arrive.

Google has quietly rolled out a whole stack of AI tools for content creation. Not one tool. A stack.

And when you connect them, you can take a single brand, start with nothing but raw files, and build an entire content engine from scratch. Blog articles. Ad creatives. Product videos. Social posts. A full website. All of it, using only Google Gemini.

I am going to walk you through the exact workflow, tool by tool, the same way I watched it happen. By the end of this email you will see why I think the content game just changed for everyone.

It Starts With A Brand Brain

The whole thing falls apart without memory. A tool that forgets your brand every time you open it is just a faster way to make generic content.

This is why the first move matters. Google rolled NotebookLM right into the Gemini app. So you build one notebook that holds everything. The hero products. The customer insights. The product images. Think of it as a brand brain that keeps expanding.

For the demo, the brand was a snack company called Healthy Crunch. Everything I describe below pulls from that one notebook.

Here is the part that clicked for me. Because the product images live inside the notebook, every asset Gemini generates later comes out on brand automatically.

You are not re-explaining your brand on every prompt. You explain it once, and the whole system remembers.

Blog Posts That Build Themselves

First task was a blog post. But not the lazy way.

You attach a content brief skill, and Gemini follows your exact brief template using the notebook as the source. Then you flip on canvas mode and ask for the full post. Out comes a complete article with proper formatting and internal linking already done.

Then it gets better. Turn on image mode and ask for a feature banner plus two supporting images. Because the product images are in the notebook, the visuals come out on brand, featuring the actual product. One click exports the whole thing to Google Docs, visuals included, ready for your team to comment and collaborate.

The same notebook also spits out a lead magnet PDF, like a lunchbox planner for parents. Or an internal one-page retail line sheet for your sales team. Less than ten minutes each.

### 100+ Claude Code hacks to ship code 10X faster

Top engineers at Anthropic and OpenAI say AI now writes 100% of their code.

If you're not using AI, you're spending 40 hours doing what they do in 4.

These [100+ Claude Code hacks](https://magic.beehiiv.com/v1/5f7ce6e3-9a71-416b-99a7-606c5f7e2447?email=bob@reisner.info&redirect_to=https%3A%2F%2Fcodenewsletter.ai%2Fforms%2F14166360-de71-46c4-8722-878d417fab5c&utm_source=beehiiv&utm_campaign=BUBDNZO6MS&redirect_delay=3&_bhiiv=opp_fd865b9f-0b56-40a4-8fe6-766f563d3caf_94e90c2e&bhcl_id=5b431f3b-052d-483e-b737-59bbc76d5f19_f672f2b6-1e60-4422-8448-ee13d3aa1099_22c35bc7-5e17-4448-baa5-49c4bf23f93e) fix that and help you ship 10x faster.

Sign up for The Code and get:

  • 100+ Claude Code hacks used by top engineers — free
  • The Code newsletter — learn the latest AI tools, tips, and skills to code faster with AI in 5 minutes a day

    • [Claim your free playbook](https://magic.beehiiv.com/v1/5f7ce6e3-9a71-416b-99a7-606c5f7e2447?email=bob@reisner.info&redirect_to=https%3A%2F%2Fcodenewsletter.ai%2Fforms%2F14166360-de71-46c4-8722-878d417fab5c&utm_source=beehiiv&utm_campaign=BUBDNZO6MS&redirect_delay=3&_bhiiv=opp_fd865b9f-0b56-40a4-8fe6-766f563d3caf_94e90c2e&bhcl_id=5b431f3b-052d-483e-b737-59bbc76d5f19_f672f2b6-1e60-4422-8448-ee13d3aa1099_22c35bc7-5e17-4448-baa5-49c4bf23f93e)

    Nano Banana Is The Quiet Star

    You already saw Nano Banana make those blog visuals. On its own it does far more.

    Pick image mode and you get pre-built styles. Hand it a few product images, pick a style, and ask for a product ad shot. The first draft is genuinely impressive even with a lazy prompt.

    But the real magic is iteration. Ask for a website hero banner. Want to remove a product in the middle? Circle it with the sketch tool and ask. Nano Banana understands the instruction and keeps going until the shot is right.

    Three things I want you to steal here:

    It does A/B testing for you. Give it a product image and ask for three text-to-image prompts with different backgrounds, moods, and styles. Pick the one you like and generate. Endless campaign angles on demand.

    It resizes creatives. A square ad that needs to become vertical for another channel takes about ten seconds. Same concept, new ratio, everything consistent. This one will save you hours.

    It never starts from a blank page. Every asset begins from a draft, not from nothing.

    Build It Once, Reuse It Forever With Gems

    Here is the catch with everything above. You have to prompt it every time. That gets old fast.

    So you build a Gem. A custom Gemini loaded with your brand and workflow once. Give it a name, write the instructions, set the default tool, and attach the knowledge.

    The example was a Studio Ad Shot Gem. Default tool set to Nano Banana. Knowledge included the product packaging dimensions, which is a small detail that makes a big difference, because now the model knows your real product size and generates more realistic images. Add a style guide. Attach a notebook.

    Now every ad shot comes out consistent, and you can share the Gem with your whole team. One Gem for studio shots. One for lifestyle images. One for campaign copy. Your team stops rewriting prompts and starts working from a standard.

    Now The Video Gets Serious

    Google has the strongest video model out there, and the new Omni model takes it further. It takes almost any input and generates almost any output, with strong physics and real-world reasoning. Product videos, reels, shorts. Most of what a marketer needs.

    Pick video mode, choose a style, drop in a few images. I asked for a product video with a female model holding the snack. Up to ten seconds inside the Gemini app.

    The character consistency floored me. It really was the same person the whole way through. And you can change the scene with a prompt while keeping everything else locked, like Nano Banana but for video.

    One rule that saves your quota. Plan the storyboard first. There is a Gem that generates an Omni video storyboard using the official prompt guide, scene by scene, and then renders a storyboard image so you can adjust before you ever spend a generation.

    Upload those storyboard frames as raw input and the final output is far more predictable.

    Omni can even reference video. Feed it a ten-second Instagram reel for pacing and editing style, hand it the product, ask for a recipe reel with real instructions, and it gets the steps right. Complex task, correct output.

    Flow Wraps A Studio Around It

    The frustration with video in the Gemini app is real. A few generations a day, no timeline, no editor, no way to connect clips. That is exactly where Google Flow comes in.

    You create a brand character once with a reference image and a prompt, give it a name and a voice, and reuse it across everything. Tag the character, tag the product, generate. Strong consistency, real-looking product, almost perfect on the first draft.

    Flow also has an agent mode. Turn it on, hand it a product, and ask for three ad variations with different camera moves and angles. The agent asks for your approval, then orchestrates all three generations for you. It follows storyboards too, and the quality almost always comes out better when it does.

    The Tool That Does It Without You Driving

    Everything so far still needs you in the driver seat, prompt by prompt. Pomelli is the one that drives itself.

    You give it your website. It pulls your brand straight off the page and sets up your business DNA. Colors, taglines, catalog, assets, all collected automatically. Healthy Crunch again for the demo, and it grabbed most of the product assets from the site on its own.

    Then you create campaigns two ways. Use a pre-built campaign it suggests, like a World Chocolate Day set with on-brand creatives ready to tweak. Or use the chat box.

    Pick a few products, give it the offer and the goal, choose the aspect ratio, and it writes a creative brief just like you would plan a real campaign. Confirm, and it generates every asset, on brand, product front and center.

    It does photoshoots from your existing product assets. It even builds a simple website from your brand context that you can publish in one click, good enough to test a campaign idea fast.

    I walked away genuinely impressed. A free tool that hands you creative inspiration and on-brand assets in minutes.

    ### Defend Your Team's Hours with Timeglass

    When a client disputes an invoice, "our timesheet says so" isn't an answer. [Timeglass](https://timeglass.ai/?utm_campaign=BUBDNZO6MS&utm_source=beehiiv&utm_medium=newsletter&utm_content=secondary1_defensible&_bhiiv=opp_163303ba-c1fa-4d3b-8aaf-d9969146944b_b6931f78&bhcl_id=0be7728d-474c-4b19-8c8c-aacf0c1e7270_f672f2b6-1e60-4422-8448-ee13d3aa1099_22c35bc7-5e17-4448-baa5-49c4bf23f93e) records what was done, for whom, with the reason attached - read from the work, not written from memory.

    [Bill with confidence](https://timeglass.ai/?utm_campaign=BUBDNZO6MS&utm_source=beehiiv&utm_medium=newsletter&utm_content=secondary1_defensible&_bhiiv=opp_163303ba-c1fa-4d3b-8aaf-d9969146944b_b6931f78&bhcl_id=0be7728d-474c-4b19-8c8c-aacf0c1e7270_f672f2b6-1e60-4422-8448-ee13d3aa1099_22c35bc7-5e17-4448-baa5-49c4bf23f93e)

    The Whole Engine In One List

    Here is the entire workflow in the order you would actually use it. Save this.

    One, NotebookLM. Build the brand brain. Load products, insights, and images once so everything after comes out on brand.

    Two, Gemini canvas plus image mode. Blog posts, lead magnets, and internal docs with visuals, exported straight to Google Docs.

    Three, Nano Banana. Ad shots, hero banners, A/B prompt variations, and instant creative resizing.

    Four, Gems. Lock your repeatable tasks into custom, shareable assistants so nobody rewrites prompts.

    Five, Omni model. Product videos and reels with strong character consistency. Storyboard first to protect your quota.

    Six, Flow. A full studio with reusable characters, a timeline, and an agent mode that orchestrates multiple video generations.

    Seven, Pomelli. The hands-off option. Reads your site, builds your brand DNA, and ships full social campaigns and even a website on its own.

    My Final Thoughts

    I keep coming back to that invoice I stared at two years ago.

    Everything I paid a team to do across days, this stack now does in an afternoon. Not at agency quality on every single asset, not yet.

    But close enough that the gap is shrinking faster than most people realize.

    Here is the real shift though, and it is not about saving money. It is about speed of testing.

    When a blog post takes ten minutes and an ad variation takes ten seconds, you stop guessing which idea works.

    You just make all of them and let the results tell you. The people who win the next two years will not be the best designers. They will be the ones who test the most ideas the fastest.

    You do not need all seven tools today. Pick one. Build a NotebookLM brand brain this week and run a single blog post through it. That one step will show you what the rest of this is capable of.

    The hard way of making content is not dead yet. But it is on the clock.

    Stay curious, talk to you tomorrow.