Hello Reader

When Jodie Woodworth found her team down a product marketing manager (PMM) in early 2026, her Brand President said no current backfill until Q4.

She had a full PMM workload, a five-person team already at capacity, and no clear answer for how any of it was going to get done.

Jodie is Head of Marketing at a 200-person Fortune 500 subsidiary a lean team operating with relative independence inside a 20,000-person parent company. She has access to pre-approved tools only, no API connectors, a Google Workspace environment and a 14-person sales team clamoring for sales enablement materials.

Jodie’s situation isn’t unusual.

Hiring freezes and holds on backfills are playing out across marketing teams right now.

A 2026 CMO Survey from Deloitte puts marketing headcount growth at 50% below the prior year and marketing budgets are now under more scrutiny than ever before.

Most marketing leaders respond by redistributing work across an already stretched team. Jodie looked at what the PMM role produced and built systems that could produce the same outputs without a headcount.

Three to six months later, inbound leads are up, the sales team has a live product knowledge base instead of a static folder, and the product team writes better documentation.

The Situation

At a B2B software company running a formal product lifecycle process, the PMM is the connective tissue between product, marketing and sales. Without a PMM, none of the knowledge transfers into positioning guides, competitive analysis, sales sheets or training materials.

It either piles up or doesn’t get done.

Jodie’s team has five people: a growth manager, a senior creative designer, a content writer, an events manager and a marketing specialist. None of them had bandwidth to absorb a PMM function on top of their own roles.

So what did she do?

The System

Jodie’s company approves Claude, ChatGPT and Gemini, but the Google suite of products (Gemini, NotebookLM) offers the most robust functionality for what she needs.

The core build is one Gemini Gem per output type. Each Gem is trained on the Pragmatic Institute go-to-market framework and fed a consistent set of inputs:

• JIRA notes from the product team
• Meeting transcripts
• Product presentations
• Brand voice documents

She built separate Gems for each deliverable: the product positioning guide, sales sheets, email sequences and competitive analysis.

The positioning Gem produces a 5 to 25-page guide per product, covering the positioning statement, product-market fit, product description, feature-benefit table, competitive breakdown, target audience, pain points, case studies and buyer and seller FAQs.

What used to take the team weeks to produce is now done in a fraction of the time. A simple product takes about two hours to set up. A complex one, with multiple integrations, layered feature sets, and more nuanced messaging, requires more iteration, but still wraps up in hours rather than weeks.

Working with Gemini

Jodie describes her design process as a joint effort with Gemini—part intuition from 20-plus years of marketing experience and part the model identifying what’s missing.

“I’ll ask it, what questions should I be including so you can find this information faster? What instructions do I give the Gem to make this process repeatable?”

She also says it's key to assign each Gem only one job. As an example, the positioning Gem does only positioning and the competitive Gem handles only competitive research. No Gem carries the entire PMM function, which keeps context windows manageable and output quality consistent across products.

There was one beneficial side effect she didn’t anticipate. The product team changed how they write JIRA notes. They noticed that vague inputs produced gaps in the positioning guide output, so they started being more specific, answering the questions the Gem needed before anyone had to ask. That behavior change wasn’t a mandate, it just emerged from seeing what the system required.

But Jodie knew organic adoption wouldn't guarantee quality at scale. To prevent the system from degrading, she locked this behavior in with a process mandate: a required update to the company’s Product Lifecycle Process (PLCP) template. By mandating a standardized, SME-approved PLCP document as the single source of project truth, she engineered true cross-departmental accountability.

The Jodie Approver

Before any draft reaches Jodie for review, it passes through a Gem her team trained on her own feedback.

They collected her written critiques over time, detailed and specific notes on what worked, what didn’t and why. Those notes became the training material for a Gem that replicates her review patterns. Drafts that clear the Jodie Approver arrive in her inbox closer to done.

“I personally love it,” she says, “because now I spend less time approving things.”

The team gets faster feedback loops and Jodie gets back review hours she was spending correcting things the Gem now catches first. And the Gem keeps improving as her feedback continues to accumulate.

Here's a copy of the Jodie Approver prompt instructions:

NotebookLM as the Brain

The second half of Jodie’s system lives in NotebookLM, which functions as the shared knowledge base for marketing, sales and customer success.

It holds 300-plus documents, all connected directly to Google Drive.

When a positioning guide updates in Drive, the notebook reflects it. When a new product recording gets added, it’s immediately available to query. No manual uploads, version control problems, or stale documentation.

Here’s how the team uses it:

Marketing

The marketing team uses it to produce ad campaigns, blog posts and ABM email sequences. The marketing specialist loads an account-specific notebook, built around brand language, priorities and known pain points, and uses it to check whether the tone and style of outbound copy.

Jodie says the result has been a noticeable increase in inbound leads over the past three to six months, with prospects arriving further down the funnel and primed for a conversation.

Events

The events manager uses it to write talk tracks and scripts customized to the ideal customer profile (ICP) attending each event. Rather than writing generic messaging and hoping it lands, he goes into the relevant product notebook and pulls positioning that matches the audience in the room.

Sales

This workflow also accelerates the entire Go-To-Market motion. By feeding the approved guides into NotebookLM, the team turned static documents into an interactive 'tutor' for the sales team.

For day-to-day product knowledge, sales reps ask questions in plain language and get an accurate answer in thirty seconds, rather than digging through a 25-page guide or waiting on someone in the Product team to answer a question.

Jodie’s team is currently building this prompt to turn the static Product Positioning Guides into active sales coaching:

What’s Next

Jodie is building the next layer now.

The plan is to connect each product notebook to a Gemini Gem that runs interactive sales training: a rep picks a customer scenario, the Gem plays the prospect and the rep works through objections using the product knowledge they’ve built. These practice conversations are intended to mirror what they’ll face in the field, with no passive reading required.

What Didn’t Go Smoothly

Sales adoption is still incomplete. Of the 14-person team, Jodie doesn’t have full participation.

“It’s fear,” she says. “They’re asking: what if it gives me the wrong answer? I don’t know how to use it.”

Even with a tool designed to be queried in plain language, the conversational interface is unfamiliar. Jodie works through it with one-on-ones, all-hands mentions and a cross-team AI chat space where anyone can ask questions or share what they’re trying.

Hallucination is also a real risk.

Jodie stress-tests Gems regularly rather than setting them and walking away. “You almost have to treat it like a person,” she says. “You have to continue to train it, coach it, give it the lessons and guidance.”

If you stop paying attention, the output drifts.

Rethinking the System

When a key person leaves under a headcount freeze, the instinct is to absorb the work and spread it across the people who are left. If you’ve ever had to do that with your team, you know it's an unpopular choice.

Jodie took a different route.

She created a live knowledge infrastructure multiple departments use. Marketing produces more content in less time, sales has product knowledge on demand, and the product team documents more carefully because they understand the downstream effects.

And while the system is proving invaluable, Jodie admits it doesn’t fully replace the role. "We will still hire another PMM to manage the program and drive our capabilities," she says. "But when we do, this system will dial them in a lot faster."

Claude for Marketing Leaders: Two Workflows in Two Hours runs on May 26. We build a competitive intelligence system and a content automation system live — every step demonstrated, then replicated.

As an AI at Work subscriber you get $50 off.

Use code AIWORKS at checkout.

$249$199

​Register now →​