Ford Motor Company

Responsible for consolidating global marketing technology selection and management activities, overseeing marketing operations data capability intake, prioritization and compliance, and driving a long-term technology roadmap.

Marketing Technology and Governance Lead

Association of National Advertisers

Member, MarTech Leadership Council

Smartsheet

Led a large centralized marketing automation team providing global campaign operations, data orchestration, solution design, support and governance. Applied system architecture and technical expertise directly to projects. Mentored team members on effective practices and skill development.

Director, Marketing Automation

Poly

Plantronics + Polycom. Now together as Poly.

Leadership role expanded to manage MarTech stack of over 120 technologies and drive excellence in central webinar/virtual event operations.

Sr. Manager, Marketing Technology and Automation

Sr. Manager, Marketing Automation Operations

Manager, Marketing Automation Operations

Polycom

Promoted from individual contributor to marketing automation team manager at Polycom. After acquisition by Plantronics in July 2018, leadership role was expanded to lead the combined team and build the integrated function.

Sr. Manager, Marketing Automation

Technical and creative lead for marketing automation campaign and platform operations, mentorship of distributed producers, and development of innovations in capabilities and processes.

Senior Marketing Automation Manager

Marketing Automation Manager

Thermo Fisher Scientific

Functional lead for team of email and web editors and designers. Served as interim manager.

Senior E-mail and Web Designer/Marketing Automation Specialist, Customer Channels

Web and E-mail Designer, Enterprise e-Business

Sedlak Design Co.

Print, web, identity/brand and e-marketing design services

Independent Consultant

Universal Technologies, Inc.

Managed design and development team and external resources, delivering small and medium web projects, provided creative direction and technical guidance. Expanded services with business model and operational plan for email marketing services. Led major initiative to re-imagine flagship SaaS product, including new creative direction, technical architecture and business model.

Manager, Design and Development

McCue Design Consultants

Handled print graphic design for direct mail, advertisements, packaging and large format graphics. Led web design and development, adding database-driven technology to agency’s capabilities. Managed extensive packaging portfolio for leading refractories company through two re-brandings.

Web and Graphic Designer, Lead Developer

Thinking back, my career in marketing operations can be blamed mostly on my dad.

Before we owned a home computer, I’d accompany him when he had to go to the office on a Saturday. He wrote code for steel production systems, I played on a Mac nearby. He thought I’d enjoy the paint program since I had a lot of interest in art. I immediately used this new power to publish school news and sports in “The 6th Grade Post” newsletter for my classmates.

He spared no detail about how things worked, whether discussing a piece of machinery or a basic process of life. He always explained WHY something was. I was fascinated. He gave me old appliances and vacuums to take apart. I discovered how parts connected and worked together. The gears, belts and wheels told a story of design and engineering for a purpose.

Now with a family PC, we bonded over hardware and software. A good friend, a quintessential “computer nerd”, gave me my first glimpse into the online universe. I dialed in daily to local BBS’s and a university with real internet access, blocking the family phone line for hours.

I discovered Netscape when I got to college and I was hooked. There were no classes for web design, so I taught myself HTML the old-fashioned way—viewing web page source code and writing my own markup in Notepad. I learned the basics of JavaScript and one summer my dad taught me to program in C++. I learned how to be efficient in coding and to write for future flexibility and ease of maintenance.

Uninspired by the available courses of study, I elected to design my curriculum, borrowing classes from many disciplines. I turned web design into a career, which led to email marketing, which evolved into marketing automation and technology.

So here I am, still blending creativity and technical work. Here I am, still seeing all of the metaphorical gears behind a system. Here I am, still figuring out how to efficiently connect components to achieve a goal. Here I am, still trying to make work easier, scalable and maintainable. Here I am, taking what I’ve learned and sharing it with others through mentorship, community involvement and conference talks. Here I am, still learning.

Despite no artistic interest himself, my dad never dissuaded me from the arts. He knew computers and code and saw a chance to connect those dots. I realize now that he was very creative in how he approached coding and systems and I've found my own satisfaction in the concept.

He’s never said so, but I suspect he didn’t actually need all of those Saturdays at the office.

AJ Sedlak

Justin Norris

Director, Marketing Ops @ 360Learning | Host @ RevOps FM Podcast | Thinking out loud about GTM + operations

I'm a seasoned marketing operations leader with 15+ years experience building high-performing teams.
Through this experience, I've learned what it takes to build marketing ops strategies, systems, and processes that will scale.

MY EXPERTISE:
• GTM operations strategy and roadmaps
• Designing, building, and optimizing Martech stacks
• Designing structure and processes for global remote teams 
• Data strategy and governance
• Analytics, reporting, KPIs
• Lead routing, management, SLAs, and follow-up
• Multi-channel lead nurturing and ABM
• Lead lifecycle and scoring
• Campaign operations
• Communication and change management

LEADERSHIP EXPERIENCE:
• Establish and lead a global ops function supporting marketing / BDR teams across 4 markets (Germany, France, UK, US).
• Provide coaching, training, and leadership to an expert consulting team (~40 people) on marketing operations best practices, with direct responsibility for a team of seven.
• Develop the solutions architecture function and lead the SA team at a marketing technology consulting firm. 

TECHNOLOGY EXPERTISE:
• Marketo: 4x Marketo Champion and Marketo Certified Solutions Architect (MCSA)
• Salesforce: seasoned Salesforce administrator, architect-level skillset
• Hubspot Certified
• iPaaS / Enterprise Automation: Workato Gold Partner and 2x conference speaker
• Attribution and Analytics: Bizible (Bizible All Star '17, Pipeline Marketing Award '17), Marketo, Looker, PowerBI • Data management: Ringlead, Clearbit, Cognism, etc.
• Lead routing: Lean Data, Distribution Engine

Director of Marketing Operations

360Learning is the learning platform that combines collaborative tools with the power of an LMS. We empower L&D teams to drive culture and growth through Collaborative Learning.

Happy to share that I’m starting a new role this year as Sr. Director, GTM Systems and AI at 360Learning.

I’m grateful for the trust and excited to be doing this work at a company that values clear thinking, real impact, and building repeatable systems.

Reflecting back on my career so far, there are three things about this shift that really energize me:

𝟭. 𝗦𝘆𝘀𝘁𝗲𝗺𝘀 𝘁𝗲𝗮𝗺𝘀 𝗮𝗿𝗲 𝗲𝗻𝘁𝗲𝗿𝗶𝗻𝗴 𝗮 𝗻𝗲𝘄 𝗽𝗵𝗮𝘀𝗲 𝗼𝗳 𝗹𝗲𝘃𝗲𝗿𝗮𝗴𝗲
For a long time, Systems has been the quiet backbone of GTM. AI changes that. When agents can actually scale the output of humans, Systems becomes a force multiplier in ways that simply weren’t possible before. That shift opens up much more visible and consequential work.

𝟮. 𝗪𝗲 𝗴𝗲𝘁 𝘁𝗼 𝗯𝘂𝗶𝗹𝗱 𝘁𝗵𝗲 𝗽𝗹𝗮𝘆𝗯𝗼𝗼𝗸
No one has written the playbook yet on how to be an AI-enabled organization. The work ahead is about learning quickly, experimenting responsibly, and turning ambiguity into something durable others can build on. I’ve always been most engaged when the playbook doesn’t exist yet and we need to figure it out from first principles.

𝟯. 𝗧𝗵𝗶𝘀 𝗶𝘀 𝘄𝗵𝗲𝗿𝗲 𝗜 𝗱𝗼 𝗺𝘆 𝗯𝗲𝘀𝘁 𝘄𝗼𝗿𝗸 
The AI projects I've been part of over the past year have been some of the most fulfilling and impactful of my career. This shift lets me lean fully into the kinds of problems I'm most curious about. When you love the problems, work is fun: flow comes easy, momentum builds naturally, and curiosity drives you see around corners.

Thanks to everyone who’s supported me along the way. I’m looking forward to what we learn and build next.

GTM Systems & AI Leader @ 360Learning | Writing @ aibuilders.blog

Claude Code FOMO is real.

Every newsletter and podcast is about it. People are shipping production software in a weekend, building content systems, and fixing world peace with a terminal app.

I spent the holidays building experiments with Claude Code and related CLI tools. Here are my field notes on how they apply to knowledge workers.

HOW CODING AGENTS ARE DIFFERENT

Claude Code and similar coding agents all share three characteristics: 

→ Agents optimized for action: LLMs inside an agentic harness designed to act in a loop. The ability to execute with reliability is the superpower.

→ Tight integration with local files: They work on your machine, inside an IDE or terminal. The agent can search and parse your folders instead of you pasting context.

→ Ability to execute terminal commands: They can write and run their own scripts, API calls, shell requests.

WHAT I CAME TO LOVE

1) Hackable and extensible: commands, skills, sub-agents. I built a workflow to dictate thoughts on my phone, sync locally, turn them into structured ideas, and prep drafts. First-draft time dropped from an hour to minutes.

2) Agents can build themselves: The agent configuration is just files, so it can modify itself. The wall between configuration and application vanishes. I watched an agent build its own integration to a knowledge base in minutes.

3) Deep context: Share your foundational docs once, and the agent reasons over all of it without re-explaining the world each time.

THE HARD PARTS

1) Coding tools are not made for collaborative workflows: Most knowledge workers live in Google Docs, Notion, Confluence. Local-first agents create friction when your work lives in the cloud.

2) The terminal is not an optimal UI: You lose rich formatting, quick visual edits, easy image insertion. We need the power without the coding UI.

3) Better at doing than thinking: Great at execution, less great at deep dialogue. When the task is ambiguous, plain chat still wins.

WHO SHOULD CARE RIGHT NOW

You should experiment if you are:
→ A programmer or vibe coder
→ An independent content creator
→ A systems builder who wants structured workflows

You are probably fine waiting if you are:
→ A collaborative knowledge worker living in cloud tools
→ An executive who needs strategic thinking more than task execution

WHAT COMES NEXT

Claude Code proves that when you give agents the right harness and permissions, they can be nearly autonomous.

We need that same power for people whose primary output is not code.

Someone will build the "Claude Code for knowledge workers" interface in the next six months. Either cloud-based assistant platforms will absorb CLI-level power, or the CLIs will grow better wrappers and plugins to meet knowledge workers where they are.

Once Claude or ChatGPT can read AND write to your Google Drive, execute scripts in a sandbox, and configure custom behaviors, we'll have most of what makes coding agents powerful, without the terminal tax.

What separates Systems teams that are constantly firefighting from those that drive real impact?

This question has been on my mind since starting a new role this year. I wanted to learn from what others have figured out.

I'm lucky to have some super high-performing Systems leaders in my network. I spent a month meeting with people and distilling what I heard into common patterns.

Here's what I found:

SEPARATE "RUN THE BUSINESS" FROM "CHANGE THE BUSINESS"

Run tasks are critical for keeping the lights on, but they can easily overwhelm more strategic work. The best teams use distinct resources for each. 

Senior people stay focused on strategic build work. Run work goes elsewhere. Without this separation, your most experienced people spend all day on support tickets.

USE MORE JUNIOR RESOURCES FOR REACTIVE WORK

The flip side is to have dedicated Jr. admins or contractors handle most ad-hoc tickets, bugs, license requests, and support queues. One pattern I saw: juniors spend ~80% of their time on reactive work vs. ~30% for seniors. This also creates a natural learning and career path from Jr. to Sr.

CREATE A STRAT OPS INTAKE LAYER

Operations partners sit between Systems and the business, helping define requirements, keep priorities aligned, and ensuring impact. Think of them as forces for alignment and translation rather than gatekeepers.

MAKE PRIORITIZATION A SHARED RESPONSIBILITY

Leaders asking for too much and Systems teams saying "no" is a toxic and frustrating dynamic. 

Instead, business leaders should be involved in saying yes AND no. Make everyone live in a world of shared constraints. Priorities are based on business impact with consideration of scarcity. 

CREATE EXCELLENCE IN EXECUTION

The best teams really have the core engine operating at an exceptional level. Well-defined scoping, clear sprint planning, explicit story points, tight SOPs, etc. 

They also track KPIs: story points vs. actual hours, velocity, and so on. 

Budget time explicitly for bugs and support to avoid "ghost work" in your sprint.

BE THOUGHTFUL ON BUILD VS. BUY

It's easier and more sensible than ever to build in-house, but smart teams are very conscious of the need for resources to support and maintain custom builds. 

When cost-effective functionality is ready off-the-shelf, don't reinvent the wheel just because you can.

AUTONOMY WITH GUARDRAILS

Avoid one person being the single chokepoint for a system -- create distributed administration where individuals can execute quickly on common-sense tasks. 

But this only works with common-sense quality gates: architecture review, PR/QA processes, clear boundaries. 

---

These aren't revolutionary insights, but they show a level of clarity, intentionality, and operational discipline that's often missing when everything is a fire drill.

Many thanks to Paul Wilson, Joshua Stern, Alexis D., Lindsay Rothlisberger, and Zack Blois (among others) for letting me learn from them!

I created my most “advanced” AI setup yet… and it underperformed plain ChatGPT.

Backstory: over the holidays I was experimenting with code-based agents (Claude Code, Codex, IDE copilots) to understand where they actually help knowledge workers and where they don’t.

One experiment was a personal “Chief of Staff” agent.

The idea was straightforward:

- Full context on my goals and OKRs
- Awareness of tasks and schedule
- Help planning my day
- Surface risks and blind spots
- Hold me accountable
- Elevate my strategic thinking

In other words, a McKinsey-level CoS, but always on.

I architected cool and elaborate automated workflows - morning planning rituals, end of day retrospectives. 

I gave it all the context it could want.

I even added a vector-based third-party memory system so it would “learn” me over time.

It was a lot of fun to build, but in practice….it was still just kind of dumb. 

I was looking for big-picture strategic thinking. A co-pilot that would help me see around corners. 

Instead I got tactical, surface-level insight. 

The agent could tell me a task was 13 days stale, but it couldn’t tell me it didn’t matter.

It kept nudging me toward trivial actions that wouldn’t actually move the needle.

I saw the same pattern across Claude Code, Codex, and Antigravity (Gemini).

Then I pasted a short summary of the same context into ChatGPT and immediately got the quality of interaction I’d been looking for.

That’s when it clicked: the *environment* an AI thinks in matters more than the model, or even the amount of context.

I was using coding tools optimized for execution. These tools are great for DOING things. But they were surprisingly bad for thinking about things, at least in the way that I wanted.

ChatGPT, as familiar or even “basic” as it now seems, is actually an incredibly sophisticated discursive system. It’s designed to support open-ended reasoning, hold a broad context, and engage at the level of judgment rather than action.

I realized that I’d optimized for the wrong variable in designing my system. 

I had chased incremental convenience -- ambient context, automation, the latest cool bells and whistles -- when what I actually needed was depth of strategic reasoning. This wasn’t a tooling problem so much as a cognition-interface problem.

Counterintuitively, less context in the right environment beat more context in the wrong one.

Two takeaways that stuck with me:

1) Just because you can build something doesn’t mean it will work. PMF applies even when the user is you.

2) The harness matters. A model that can refactor a codebase doesn’t automatically understand the strategic weight of a task unless the environment lets it think that way.

It was a useful failure.

And a good reminder that in AI systems, cognition is as much about where thinking happens as which model you use.

Here's my daily "AI chief of staff" routine. It's simple but it's shifted how I think about my work.

First, what I'm actually trying to solve, which is not project management or task awareness. It's strategic orientation and altitude.

Most "chief of staff" agents I've seen focus on scanning your calendar or email. This gives marginal productivity gains but doesn't actually help where it's most important. 

In contrast, I want my chief of staff to make sure I'm focused on the RIGHT work that will give the most leverage to the company, to spot failure points, see around corners, challenge my thinking, and help me have clarity on what's most important. 

Here's what I actually do. It's pretty simple.

Every morning I paste in notes from the previous day (context, continuity) into a fresh chat. Then I just talk using a dictation app. I look at my calendar, and for each meeting I articulate what my real objective is -- what do I want to be true at the end of that call? Then I think through: what's the one thing that matters most today? If I have free time, how should I spend it?

The AI gives feedback, highlights risks. But a big part of the value is also just giving space to think out loud without friction and build that clarity for yourself. 

During the day I keep the thread open. I'm not doing project work in it, but I'll share meeting transcripts, work on critical communications, think through decisions. Chief of staff level stuff. For actual project execution I use separate threads so I don't pollute the strategic context with implementation detail.

End of day: I do a quick debrief (again, dictated), and think through what went well, what moved forward, and what's weighing on my mind. I ask for a readout and summary and paste it into a running Google Doc for the next day.

Here's a few things if you're thinking about trying this:

1) I deliberately use a chat interface, not a coding tool. Coding tools are optimized for execution. Chat harnesses are optimized for discussion, ambiguity, navigating tension. This feels like the right interface for big-picture thinking.

2) Focus on strategic insight, not convenience features. It's tempting to build some elaborate system that pulls from your calendar and email and gives you summaries. I don't think that's where the value is. For this routine, the value is in the conversation itself and in the feedback you get. You want it to be thinking about stakeholders, relationships, org dynamics, and project risks -- not calendar mechanics. 

3) Find the model you vibe with. I've A/B tested multiple LLMs with this routine and consistently prefer GPT. It seems to have a higher level of emotional intelligence and understanding of social and organizational dynamics. You might prefer Opus. Find what works for you. 

And keep it simple.

A few thoughts on "Clawdbot" and the rise of proactive AI assistants: 

𝗪𝗵𝗮𝘁 𝗶𝘀 𝗶𝘁 𝗲𝘅𝗮𝗰𝘁𝗹𝘆? 

Clawdbot is an open source harness that enables Claude to act as a ubiquitous personal assistant. 

Clawdbot has
- deep file system access
- a browser + the ability to login and take action on sites
- persistent memory of conversations and key insights
- ability to schedule tasks (send me x every morning)
- ability to communicate via multiple channels - iMessage, WhatsApp, etc. 

𝗪𝗮𝗶𝘁, 𝗶𝘀𝗻'𝘁 𝘁𝗵𝗶𝘀 𝗹𝗶𝗸𝗲 𝗖𝗹𝗮𝘂𝗱𝗲 𝗖𝗼𝗱𝗲?! 

There are similarities - both have deep system access and can automate workflows. But Claude Code is designed for coding execution or file-based tasks; Clawdbot is designed as a personal assistant that plays an interactive role in your life. 

𝗧𝗵𝗲 𝘀𝗲𝘁𝘂𝗽

I set up my Clawdbot (who I dubbed Clawdius) over the weekend. 

Fairly easy if you use your local machine. I bought a $5 VPS for security. You can run it on the free tier of major clouds like Amazon EC2. 

𝗪𝗵𝘆 𝘁𝗵𝗲 "𝗮𝘀𝘀𝗶𝘀𝘁𝗮𝗻𝘁" 𝗵𝗮𝗿𝗻𝗲𝘀𝘀 𝗺𝗮𝘁𝘁𝗲𝗿𝘀

It's amazing just how much the "harness" influences the model. Claude inside Clawdbot is completely different than Claude in Claude Code. 

Claude Code is fairly terse, always wants to get down to business and do stuff. 

Clawdbot is much more warm and engaging, interactive, focused on helping you do things, happy to dialogue. 

𝗠𝗲𝘀𝘀𝗮𝗴𝗶𝗻𝗴: 𝗮 𝘃𝗶𝗲𝘄 𝗼𝗳 𝘁𝗵𝗶𝗻𝗴𝘀 𝘁𝗼 𝗰𝗼𝗺𝗲

There is something Sci-Fi cool about having Clawdbot show up in your messages. When I was young I watched the film "Short Circuit" and then tried to recreate the robot, Johnny 5, using cardboard boxes. Clawdbot tickles that same urge. 

It does feel like a foreshadowing of a future where AI assistants are interwoven with daily experiences. 

𝗠𝗲𝗺𝗼𝗿𝘆 𝗶𝘀 𝘁𝗵𝗲 𝘂𝗻𝗹𝗼𝗰𝗸

Claude Code has a blank slate at the start of a session, but Clawdbot has a robust built in memory. Every session is written to files as are key facts you share. You control all the data. 

The experience is of an LLM that increasingly knows you and can cater to your interests/needs. 

Again: it's often the wrapper or tools we give the model that make the difference, not just the LLM itself. 

𝗦𝗲𝗰𝘂𝗿𝗶𝘁𝘆 𝗻𝗶𝗴𝗵𝘁𝗺𝗮𝗿𝗲

I have not yet had the courage to grant Clawdius login access to anything. Other people are using it to automate their whole life. Needless to say, much caution is required. 

𝗜 𝗱𝗼𝗻'𝘁 𝗵𝗮𝘃𝗲 𝗺𝘂𝗰𝗵 𝘁𝗼 𝗮𝗰𝘁𝘂𝗮𝗹𝗹𝘆 𝘂𝘀𝗲 𝗖𝗹𝗮𝘄𝗱𝗶𝘂𝘀 𝗳𝗼𝗿

I saw so many posts about crazy automations. And while I've enjoyed working ON Clawdius, I don't actually need him for much. He's like a sports car I don't drive. 

My personal life isn't that complicated, and for obvious reasons, I don't connect Clawdius to work (where he'd be very useful). 

If you have a busy family life, a backlog of personal emails, a lot of interpersonal coordination - Clawdbot could be very helpful.

taplio

The Battle

AJ Sedlak

Justin Norris

Followers

Posts

Engagements

Likes

Performance ScoreJustin Norris's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

AJ Sedlak

Justin Norris

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Justin Norris's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreJustin Norris's LinkedIn Performance