Marc Sirkin

Join Marc Sirkin for a hands-on AI marketing strategy training to learn how to configure your AI environment and build the analysis and decision workflows most marketing teams have never attempted.

Knak's Marketing Production in the Age of AI survey found that 70% of enterprise marketing teams have deployed AI, but only 29% describe their adoption as advanced. Meanwhile 82% still spend at least half their time on production instead of strategy, and 85% missed a campaign launch date last year. Faster drafts did not fix any of that, because drafting was never the slow part.

This Master Class is about everything else: the research, analysis, prioritization, and planning work that decides whether the content was worth making in the first place.

You will leave with a working Search Console workflow that turns your own search data into a defensible content plan, plus templates for several more.

No coding. No API keys. No sponsor and no product pitch.

Online December 9, 2026 | 11:00am - 3:30pm ET | $199

• How these systems actually work, and how they fail: Context windows, non-determinism, and the four failure modes that quietly ruin marketing work: confabulation, sycophancy, silent truncation, and generic output. Demonstrated live, on purpose.

• Context engineering, not prompt engineering: Output contracts that define the answer's shape before you ask. A mediocre prompt with strong context beats a brilliant prompt with none.

• Configuring an AI environment that survives past the demo: The five primitives every major assistant shares, mapped across Copilot, Claude, ChatGPT, and Gemini. Built live, including what a locked-down tenant will and won't allow.

• Search Console to a content plan: Striking-distance queries, cannibalization, intent gaps, clustered by job-to-be-done into a ranked plan you can defend. Built on your own export, live. This is AI deciding what to write, not writing it.

• Diagnosing performance: Pipeline leakage, stage decay, channel quality. Plus competitive analysis with every claim traced to a source, and a dissent pass that red-teams your own conclusions.

• Solo to team: Turning a working session into a shared asset: versioned context, review gates, handoffs that don't degrade output. The step most marketers skip.

• Agents... what they are, what they aren't, where this goes: Live demos of agents doing real marketing work, then the honest map of cost, failure modes, and what exists in which tool as of December 2026.

• Open Q&A + additional assets: The synthetic dataset, workflow templates, configuration checklist, and a 30-day follow-up series.

This Master Class is for marketers who already use AI most days and have hit a ceiling somewhere around drafts, images, and summaries. Marketing operations, demand generation, content strategy, growth, and marketing leadership will all find something to take home.

We start with foundations, and they are not table stakes. Most people using AI daily have never been taught how these systems fail, and it is the single largest cause of bad output and misplaced trust. If you already know what an output contract is, why non-determinism matters, and how context windows behave under load, this session is not for you. If you have never heard those terms, or you have heard them and could not explain them to a colleague, you are exactly who this is built for.

No coding is required. No API keys, no developer tools, and nothing you cannot run from a browser.

• Read Why AI hasn't solved marketing's time problem: https://martech.org/why-ai-hasnt-solved-marketings-time-problem/

• Download Knak's Marketing Production in the Age of AI report and skim the production benchmarks: https://knak.com/state-of-marketing-production/

• Review a prompting fundamentals guide so we can move straight past the basics: https://docs.claude.com/en/docs/build-with-claude/prompt-engineering/overview

• Export your Search Console data before the session. Queries and Pages, last 12 months, CSV. We will build on it live. Sample data is provided if you cannot get an export, but this workflow is far more useful with your own numbers.

• Test one thing: Upload any file to whichever AI assistant you plan to use at work. Many enterprise deployments block this, and it is better to find out now than at 11:15 am on December 9.

Here’s a look at what you’ll cover when. If you can't attend live, on-demand access is included for free. (Exact times are subject to change.)

11:00am – 12:30pm ET | AI Foundations

12:30pm - 12:45pm ET | Break

12:45pm - 2:15pm ET | Building Workflows

2:15pm - 2:30pm ET | Break

2:30pm -3:30pm ET | Advanced Session