How I work.

The thesis

Most marketing functions are not bottlenecked on strategy. They are bottlenecked on the operational infrastructure that supports it: manual reporting, fragmented attribution, content production that cannot scale, and a long list of point tools that no one fully connects. My work focuses on rebuilding that infrastructure so the strategy can land. Done well, this materially reduces the cost and time required to operate a modern marketing function while improving measurement and output quality.

Areas of focus

1. Attribution and reporting

I build closed-loop attribution from ad impression to revenue, with the warehouse and CRM agreeing on one definition for each lifecycle stage. The deliverable is a single weekly report the executive team can act on without an analyst rebuilding it every Monday.

2. AI-augmented content production

I architect content pipelines that combine large language models, MCP-based data access, and existing marketing tools (Make.com, HubSpot, Buffer, scheduling platforms) into a single automated workflow. The objective is consistent, brand-aligned output at meaningfully higher volume than a manual process, with human review at the points where it adds value rather than at every step.

3. CRM and lifecycle automation

Modern marketing teams underuse their CRM. I rebuild lifecycle stages, scoring, and workflow automation inside HubSpot or Salesforce so that segmentation and handoff happen automatically and consistently. I also build and maintain Model Context Protocol servers (see my open-source hubspot-mcp) so that AI assistants can query and update CRM records directly under appropriate permissions.

4. Paid media operations

On the paid side, I focus on the operating infrastructure around media buying: campaign architecture, creative rotation cadences, negative-keyword harvesting, and measurement integration with the broader attribution stack. Channels I have managed at scale include Meta Ads, Google Ads, Amazon Advertising, and TikTok Ads.

How I structure an engagement

My typical first ninety days in a new role follow a consistent pattern:

• Weeks 1–2 — discovery. I conduct interviews with sales, customer success, and recent customers; map every paid channel, lifecycle stage, and report; identify the single highest-leverage bottleneck.
• Weeks 3–6 — infrastructure rebuild. Unified attribution, reconciled lifecycle stages, automation in place of manual handoff. By the end of this phase the executive team receives a consistent weekly report.
• Weeks 7–12 — operating cadence and content. Weekly leadership review installed, AI-assisted content pipeline brought online, paid media reorganized against the new attribution model, and the first hire or contractor brought on to extend the operating capacity.

Specific deliverables I commit to

• A documented attribution model that ties paid spend to closed revenue, by day 60.
• A weekly one-page leadership report covering pipeline, spend efficiency, and content velocity, by day 30.
• One AI-augmented content or operations pipeline live in production by day 75.
• An annual marketing plan with quantified targets and named owners by day 90.

How I think about budget

I treat the marketing budget the way an engineering team treats infrastructure spend: instrumented, reviewed weekly, and owned in a single place. The first ninety days are not about expanding budget; they are about earning the right to do so by demonstrating measurable improvement in a number that matters to the business.

What this is and is not

The specifics flex based on the company. A developer-tools company will likely lean toward product-led growth and technical content; a vertical SaaS will lean toward events and customer stories; an enterprise software company will lean toward account-based marketing and sales enablement. The constants are: measure carefully, automate what is repeatable, and ship the operating cadence the executive team trusts.

If you are evaluating marketing engineering hires or considering a project, I’d be glad to discuss your specific situation. Get in touch or email me directly.