The End of Linear Marketing

We’ve noticed a recent change in how we approach our marketing processes. What was once a linear process of brief → create → review → approve → launch → measure → repeat, is now more like a fluid cycle.

Our campaigns no longer have a “start” and a “finish.” Our assets are no longer created as “finals.” Everything is in a loop of continuous, always-on optimization based on data, iteration, and refinement.

Until very recently, we achieved (or more like, “created”) these processes by stitching together various tools, agentic workflows, AIs, and so on. We’ve made oddly close relationships with n8n and other pixel-based duct tapes.

This week alone, we are already seeing a pattern forming. What we have been saying for months is becoming true: AI moves to the infrastructure layer. We won’t need pixelized duct tape anymore. Everything will be run directly inside our chosen AI tool.

That means our job as marketers won’t be to move work forward anymore. It will be simple: keep the system in motion. Always-on, never sleeping, constantly refining.

— Peter & Torsten

Before we jump in, a word from this week’s sponsor:

Every leader hits the same wall: too many priorities, not enough bandwidth. Wing clears that wall with a full-time virtual assistant who runs the drag layer so you lead, not chase.

This isn’t gig work. It’s dedicated support that shows up every day and allows founders to delegate without losing control.

Clear your plate

OpenAI has just released Sora 2, its powerful text-to-video model that generates startlingly realistic videos with synchronized audio from a single prompt. It can handle complex scenes, multiple shots, and maintain character consistency—a giant leap forward.

The headliner feature is "Cameos," which allows verified users to drop their own faces and voices into AI-generated scenes. OpenAI also released a free, TikTok-style social video app alongside the model. It allows anyone to create, remix, and share AI-generated videos in a "Vibes-based" feed.

The app reportedly shot into the top 3 of the App Store almost immediately, proving people are hungry for dedicated AI-native platforms. Of course, this throws gasoline on the fire of deepfake concerns. The creative power is immense, but the potential for misuse is equally great. OpenAI says it's building in safety measures, but policing a platform this powerful will be its biggest test yet.

Your video production budget is about to become significantly more flexible. This significantly reduces the cost and time required to create high-quality video, allowing your team to focus on pure ideation rather than painstaking execution.

The "Cameos" feature unlocks personalized video marketing on a scale that was once considered sci-fi a year ago. And the new app? It's a brand new channel to master.

Anthropic just released Claude Sonnet 4.5, its new flagship model, and it's a beast at complex reasoning and coding. For marketers, the big news is their writing skills. The model is way less "sycophantic," which means it's less likely to agree with you if your premise is wrong, making it a much more reliable assistant for fact-based work.

The new Claude is also a marathoner, able to stay focused on a single complex task for over 30 hours without losing the plot. It's a sharper brainstorming partner and a much more efficient writer for first drafts of briefs, blogs, and ad copy. To say the least, Sonnet 4.5 is our default model these days.

It's a huge step up, but it's not a magic solution. You still need a human in the loop with good judgment and sharp prompting skills. It gets you a better first draft, not a finished product.

Better first drafts mean less time and money spent on content creation, period. Claude Sonnet 4.5 is a more dependable workhorse for synthesizing market research or summarizing customer feedback, freeing your team from manual analysis.

As an ideation partner, it's better at challenging assumptions and generating novel campaign angles.

Meta announced that starting in December, it will use data from your conversations with its AI chatbots for ad targeting. For users outside of Europe, there’s no opt-out.

Let's be clear: this is a whole new ballgame for advertisers. It creates a direct line from a user's stated intent to your ad campaign. This isn't like targeting based on clicks or demographics. Meta will be analyzing the actual words and sentiment from conversations to infer purchase intent.

If a user asks an AI for "tips on training for a marathon," they could instantly land in an audience for running shoes and apparel. The power here is undeniable, but the controversy is massive. This move walks right up to the "creepy" line and raises significant privacy concerns that are sure to attract scrutiny from users and regulators. Brands that jump on this will need to tread very, very carefully.

This could be the highest-intent ad signal ever created, potentially supercharging your ROI on Meta's platforms. You can reach customers the moment they articulate a need.

However, the reputational risk is just as big. The brand damage from being seen as invasive could easily wipe out any performance gains.

Microsoft is making moves, rolling out "Agent Mode" for Microsoft 365 Copilot. It's an autonomous agent you can hand entire projects to, like creating a market research report or generating a quarterly sales analysis from a simple prompt.

This agent works across the Office suite without your supervision. It can pull data into Excel, write a summary in Word, and create the deck in PowerPoint, all from a single command. With "Copilot Connectors," it can even pull in data from outside platforms like Google Drive, breaking down the walls between your data silos.

This is one of the first mainstream examples of a true AI agent for knowledge work. It also opens up a Pandora's box of data governance challenges. Granting an AI system free rein over sensitive company data requires a robust plan for permissions and oversight.

"Agent Mode" is designed to automate the tedious, time-consuming work that slows down your marketing, sales, and analyst teams. It aims to kill the "grunt work" of reporting, freeing up your sharpest minds to focus on strategy. The role of your team is about to shift from doing the work to delegating it effectively.

Perplexity launched its AI-native browser, Comet, and then made the premium version free for everyone. Unlike Chrome, Comet doesn't just give you a list of links. It understands what you want and gives you a direct, synthesized answer. It’s real magic, though, is a background assistant that can run tasks for you—like drafting emails or booking flights—across hundreds of apps.

The browser is designed to be an intelligent layer over the web, not a replacement for your workflow. It's built to kill the thousand tiny cuts of digital busywork that drain your team's day.

We've been using Comet for several months, and it has become our default browser. That's saying something, because browsers are sticky. This week alone, it scraped a messy FAQ page into clean text, “translated” a landing page from US to UK English, and found a specific quote in a two-hour podcast. It's a masterclass in reducing friction.

Comet is a straight-up productivity booster for any knowledge worker. It streamlines market research, competitive tracking, and all the other little web tasks that add up to hours of wasted time. It’s a glimpse into a future where you don't just browse the web—you delegate to it.

Small Headlines. Big Shifts.

» There's a New Image Model King: Tencent's HunyuanImage 3.0 just took the top spot on the text-to-image leaderboard, dethroning Google's Nano Banana. It’s open-source and gives creative teams another world-class option for asset creation.

» Nano Banana Is Open for Business: Speaking of which, Google's Gemini 2.5 Flash Image model (Nano Banana) is now generally available for production use. It’s priced at $0.039 per image, so teams can now bake it into their workflows without waiting in line.

» Claude Is Now Your Slackbot: Anthropic released a native Claude integration for Slack, letting you summarize channels, ask questions, and brainstorm right where your team works. It's designed to kill context switching and speed up your internal thought process.

» AI Answers Get Grounded: AI tools like Gemini can now pull from Google's Data Commons, a huge library of public statistics. This means that AI-assisted market research can be grounded in verifiable data from sources such as the UN, making your reports much more reliable.

AI just crossed from tool to teammate.
Benchmarks prove it performs like an expert—only faster. Agents now manage projects, analyze data, and even shape campaigns.

The real skill now isn’t execution. It’s orchestration.
That’s precisely why we built AI Ready CMO Pro—to help marketers rebuild their workflows for this new reality.

Get AI Ready CMO Pro.

Until next time,
Peter & Torsten

PS - Quick check — where are you on your AI journey?

PPS - Know a marketing leader who should be reading this? Refer them and earn rewards.