👋 Greetings earthlings,

Emma here, co-founder of Spark AI. This edition is a big one. Google dropped a model that proved they’ve still got it - and we break down how you can deploy it in your teams with step by step use cases to experiment with. McKinsey is quietly rewriting what a consulting firm looks like (lots to learn for Agencies), and a landmark UK court case on AI and copyright that… didn't actually settle much at all. And in Spark AI news Jules gets interviewed in Design Week about his book - and we are sharing some exclusive practical tips from it plus a Q&A webinar with him on 16th December to sign up to (scroll to bottom for that!).

Let's get into it.

• 1. Google just leapfrogged everyone

  • Google Workspace Flows & Gemini in your tools

  • Visual creation: Nano Banana Pro in Adobe & Gemini

  • NotebookLM & Deep Research

• 2. What we can learn from McKinsey

• 3. AI and copyright ruling doesn't provide much clarity

• 4. Our co-founder Jules is interviewed by Design Week

• 5. Try this today: The AI Strategy Canvas

• 6. Join us for a LinkedIn Live Q&A with Jules

1. Google just leapfrogged everyone
(+ new use cases to try)

Google launched Gemini 3 on 18th November. Then in the past few days, they’ve rolled out further upgrades that quietly turn Gemini into an operating system for work (see this great behind-the-scenes look in The Wall Street Journal). For Google workspace users this is big news (for OpenAI not so great) - but what does this unlock for how my team actually works?

If you are a Microsoft user and don’t want to read about Google, we wrote an article about the Copilot advancements a couple of weeks ago - check it out here)

Google’s upgrades line up across three pillars of agency work:

• How your team works day to day: Workspace Flows & Gemini in Workspace

• How you create visuals and concepts: Nano Banana Pro

• How you research and learn: NotebookLM + Deep Research

This piece is about those three pillars, and how to turn them into practical capability inside a creative or brand team.

Google Workspace Flows & Gemini in your tools

Google has moved beyond “help me write this email” and into “help me run this whole process.”

Google Workspace Flows is a new layer that lets you automate multi-step workflows across Gmail, Docs, Sheets, Drive, Calendar and Chat, with Gemini acting as the brain that reads context and decides what to do next. Think a simple Make or Zapier native inside your Google Workspace (Read more on the Google Workspace blog).

Instead of old-school rules like “if email from X then add row in Sheet,” you can now have Gemini as a thinking layer in multi-step flows. Here are three use cases we are going to try (still to test them - if you get there first let me know how its working!)

Example 1: Be on the front foot with new business enquiries

Turn messy inbound emails into structured intake docs and pre-drafted replies, instead of reinventing the wheel in the inbox every time. When a new enquiry lands in your shared inbox:

1. A Flow is triggered by an email to something like [email protected].

2. Gemini reads the email and pulls out details such as:

   • Company name

   • Contact information

   • Budget or timing if mentioned

   • Project type (brand campaign, website, social retainer, etc.)

3. The Flow writes that into a “New enquiries” Google Sheet or a templated briefing Doc.

4. Gemini drafts a polite reply that:

   • Recaps the ask in plain language

   • Confirms key details

   • Suggests a couple of call slots for an initial conversation

   • Plays the email through a Client Research gem and a Briefing Assistant gem

   • Sends an email to your CS team with background on the client and the questions they need to ask on the call

Someone on the team gives it a quick sense-check, tweaks tone if needed, and sends.

Example 2: Standardise weekly status reports from the docs you already use

An idea to use Flows and Gemini to turn the documents you already update into consistent weekly status reports, rather than manually rewriting the same story at the end of each week. Every Friday afternoon:

1. A Flow runs on a schedule.

2. It grabs links to the docs you already maintain, for example:

   • The live campaign task Sheet

   • A performance Doc or Slides deck

   • A call transcript of your project team meeting

3. Gemini:

   • Summarises progress vs plan

   • Highlights wins, risks, blockers and key decisions

4. The Flow:

   • Drops that into a status report template in Docs

   • Drafts a client-facing email with the main bullets for you to review and send

No new tools, no extra admin. You are simply standardising the end result that already lives across Sheets, Docs and decks.

Example 3: Auto-structure chaotic client briefs into your standard format

Here’s an idea to normalise messy client briefs into your house format automatically, so the team is not hunting through paragraphs to find the objective, audience and mandatories. When a client sends a “brief” in whatever format they like (rambling email, half-baked doc, copied WhatsApp thread):

1. A Flow is triggered when a Doc is moved into an “Incoming briefs” folder in Drive.

2. Gemini:

   • Summarises the ask

   • Extracts the key fields you care about, for example:

     • Objective

     • Audience

     • Channels

     • Budget

     • Timelines

     • Mandatories and constraints

3. The Flow:

   • Fills a standard Google Doc “Client Brief” template with those fields

   • Adds a row to a “Briefs” tracker Sheet with owner and due date

   • Posts a short summary and link into your team’s Chat space

Every brief arrives in a format your creatives and producers recognise, even if the client sent it in chaos.

Visual creation: Nano Banana Pro in Adobe & Gemini

On the creative side, Google’s new Gemini 3 Pro Image model, affectionately dubbed Nano Banana Pro, is one of the first image models that genuinely gives you production-ready control, rather than just ideation or prototyping (read more in this Google dev post or Gemini 3 overview).

The standout upgrades:

1. Text in images that doesn’t fall apart
   Earlier models were notorious for mangled typography. Nano Banana Pro produces clean, legible text in multiple languages, which means you can finally use it confidently (still beware of IP issues though!) for:

   • Social and OOH mock-ups

   • Pitch visuals and concepts with real copy

   • Quick explorations of type-led ideas

2. Multi-image blending with consistency
   It can combine up to 10 reference images and maintain consistent characters and style, which is useful for:

   • Keeping brand elements coherent across variations

   • Generating series-based assets (campaign sets, product families, character-led concepts)

   • Turning rough sketches and reference boards into more polished visuals

3. Grounded, factual visuals
   Because it inherits Gemini 3’s reasoning and world knowledge, it’s better at generating infographics and explanatory visuals that align with real-world details, not vague “vibes” (this is a nice summary).

I asked Nano Banana Pro to interpret this article as a frame from a stop motion film using clay models. I spent 2 minutes on it but this still demonstrates the text quality, image consistency and factual visuals all in one.

You don’t have to change your stack to use it:

• In Weavy (which we wrote about last week), it’s right there alongside all the other image models. Each run costs 15 credits.

• In Adobe Firefly, Nano Banana Pro now powers Text to Image and Firefly Boards, alongside other partner models (see the Adobe partner page).

• In Photoshop, it sits behind Generative Fill, so your existing layer and mask workflows simply get sharper tools.

Adobe is also offering unlimited generations with Firefly and partner models, including Nano Banana Pro, for Creative Cloud Pro and Firefly plan subscribers until today, 1 December. Get experimenting quick smart!

On the trust and safety side:

• Google continues to embed SynthID watermarks in generated images - we featured Synth ID in our What’s New in Creative AI podcast back in May

• You can now upload an image to Gemini and ask if it was generated by Google AI, and it will check for that watermark and tell you.

NotebookLM & Deep Research

The biggest shift for strategists and planners is happening in Google’s research and learning tools.

NotebookLM started life as a “chat with your documents” tool where you could also create fun podcasts. In its latest update, Google has turned it into a much more capable research companion (Here is Google’s update). I’m a huge fan of NotebookLM. We build a Notebook for each client (great for onboarding new team members), each course we teach, research topics, event planning. And personally I find the audio overviews and explainer video outputs incredibly helpful to consume the content a different way that suits my dyslexia.

The key updates are:

• Deep Research
  You ask a complex question, and right inside NotebookLM you can:

  • Builds a research plan

  • Roam the web to find relevant sources

  • Produces a structured, cited report

  • Lets you drop both the report and sources straight into your notebook

• Broader file support
  You can now pull in:

  • Google Docs, Sheets and Drive links

  • Word documents (yes,.docx files in Google!)

  • Images and other media

• Learning tools built
  Recent updates add things like flashcards, mind maps, quizzes and richer overviews, so you can turn a messy pile of sources into structured learning material rather than just another long report.

What’s the key message here to take into 2026?

For those of you using ChatGPT or Copilot, don’t feel you are missing out. The AI arms race is well underway. All the models are getting better all the time, and updates to your platforms will be coming in the next few months.

Your edge in 2026 won’t be “we chose the right AI model”. It will be:

• “We built a team that can adapt and harness new capabilities as they change”

• “We encoded our best ways of working into flows and systems, not just people’s heads”

• “We turned AI from a personal productivity tool into a shared capability across strategy, creative and delivery”

That’s the model-agnostic mindset we help our clients cultivate at Spark. The specific tools will keep evolving. Your team’s fluency is what compounds.

👉 Spark AI Accelerator

2. McKinsey is rewriting its own business model - should you?

If you want to understand where professional services are heading, watch what McKinsey does.

It’s been reported that roughly 25% of McKinsey's work is now delivered through "outcomes-based arrangements" rather than traditional consulting fees. They're fundamentally restructuring how they charge for value.

Their internal tools tell the story. "Lilli," their RAG system, handles 500,000+ monthly prompts and reportedly saves 50,000 consultant hours. "One-Click" helps generate presentation decks and is now used by 75% of their 43,000 employees monthly. They can now deploy operational AI agents in under an hour.

And why this matters for agencies?

If the world's most prestigious consultancy is restructuring around AI, the acceptability of this for agencies won’t be far behind.

This is exactly why we focus on this in the AI for Leaders stream of our AI Accelerator: how to think about pricing, positioning and business model when AI changes the economics of delivery. In a way thats unique to your business. Because the agencies that figure this out first will have a significant advantage.

3. The copyright ruling that didn't answer the question

Have AI model builders infringed creator’s copyright? This is a legal and ethical question that goes to the very heart of the Generative AI revolution. Last month the UK High Court finally delivered its verdict in the high stakes Getty Images vs Stability AI case, the first of several large copyright lawsuits against AI model makers to reach a verdict. Headlines called it a "landmark ruling" and a "win for AI." But if you actually read what happened, the most striking thing is what the court didn't decide.

Getty accused Stability AI of scraping and copying millions of its images to train Stable Diffusion. But Getty had to withdraw its primary copyright claim during the trial because there was no evidence the training happened in the UK. That left the court ruling on much narrower grounds.

The judge found that Stability's AI model is not an "infringing copy" because it doesn't store or reproduce the original images. Some limited trademark infringement was found related to Getty watermarks appearing in generated images. But the core question everyone wanted answered remains unanswered: Is training AI on copyrighted material actually infringement?

The judge acknowledged the "very real societal importance" of balancing the creative industries against the AI industry. But she could only rule on the diminished case in front of her.

What this means for agencies:

This is not a green light. And it's not a definitive loss for creators either. The legal landscape remains genuinely uncertain.

The UK government is currently consulting on whether to introduce a "text and data mining exception" that would allow copyrighted works to be used for AI training unless rights holders opt out. However opting out is logistically impossibe for many artists, whose work sits on client websites and platforms (think about your own work). Artists including Elton John, Kate Bush, Dua Lipa and Kazuo Ishiguro are lobbying for stronger protection. Tech companies want broader access.

For now, if you're using AI tools in client work, the same advice applies: understand your supply chain, document your processes, and don't assume the law is settled. Because it isn't.

4. Our co-founder Jules shares his insights with Design Week

Big news from Spark HQ. Jules's book Shift - AI for Agencies launched this week and immediately hit #1 on Amazon in the AI, Sales & Marketing, and Advertising categories. Design Week picked up the story, featuring Jules discussing why he wrote it and what agencies need to focus on right now.

The timing feels right. After a year of running AI Accelerator programmes with agencies of all shapes and sizes, Jules wanted to make Spark's frameworks accessible to everyone. The book distils what we've learned about what actually works when creative businesses try to adopt AI seriously.

It's not a tools guide, it’s a strategic playbook for agency leaders asking harder questions: How do we protect our margins? How do we lead our teams through this? How do we use AI to become more distinctive, not more generic?

If you've been following Spark Intelligence, you'll recognise some of the thinking. But the book goes deeper into the frameworks, the case studies, and the practical steps. It’s our Accelerator programme in a book.

👉 Read the Design Week feature 

5. Try this today: The AI Strategy Canvas

Exercise: Complete Your Personal AI Strategy Canvas (25 minutes)

The problem: You're overwhelmed by AI possibilities and don't know where to start. Every tool promises to transform your agency, but you can't do everything at once.

The solution: The AI Strategy Canvas – a practical framework for identifying opportunities that align with your agency's core strengths.

How to do it:

1. Draw a simple table with your key business functions as columns (Client Services, Research & Strategy, Creative, Production, Support Services)

2. Create three rows: Automation (AI takes over tasks), Augmentation (AI enhances human work), Innovation (AI enables new services)

3. For each function, identify 2-3 opportunities in each category:

   • Automation: Which repetitive tasks could AI handle? (Research synthesis, meeting notes, basic asset variations)

   • Augmentation: Where could AI make your core work better? (Exploring more creative territories, deeper analysis, faster iteration)

   • Innovation: What previously impossible services could you now offer? (AI-powered brand compliance, dynamic content systems, proprietary AI assistants)

4. Review your canvas and ask:

   • Which services face commoditisation risk if we don't act?

   • Where can we enhance what makes us special?

   • What new opportunities align with our agency's DNA?

Why this works: This isn't about becoming an "AI agency" – it's about becoming an even better version of the agency you already are. The canvas reveals where AI threatens your business model, where it enhances your existing strengths, and where it creates genuine competitive advantage.

Most agencies that successfully implement AI start here. Those that don't often chase shiny tools that solve problems they don't have.

Document what you discover. Share it with two colleagues.

That's how you move from reading about AI to actually using it.

Shift – AI for Agencies contains 50+ exercises like these. If you've read it and found it valuable, a quick Amazon review helps other agency leaders discover it. Even a line or two makes a difference.

👉 Shift - AI for Agencies on Amazon or Waterstones or Hatchards (and all other good booksellers)

6. Join us for a LinkedIn Live Q&A with Jules

We're hosting a LinkedIn Live where Jules will unpack what’s behind his book Shift - AI for Agencies:

• The patterns he's seen in agencies that are genuinely moving forward with AI

• Why most agency AI initiatives stall (and what to do instead)

• The questions agency leaders should be asking right now

• Live Q&A on anything from the book or your own AI challenges

Whether you've read the book or not, this is a chance to dig into the strategic side of AI adoption with someone who's been in the trenches with dozens of agencies this year.

📅 16th December 🕐 11am 📍 LinkedIn Live

👉 Register here

Can't make it live? Register anyway and we'll send you the recording.

Until next time

That’s all for today. See you next time!

Emma 

Co-founder and CEO of Spark AI

About Spark AI

Spark AI empowers creative and brand leaders turn AI curiosity into confidence through structured training and business transformation.

We have worked with 60+ agencies running AI Fundamentals workshops and AI Accelerator programmes based on our # 1 bestselling book Shift - AI for Agencies.

Trusted by Oxford University Saïd Business School and backed by Innovate UK.

👉 Find out more 

Not a subscriber yet?