I direct marketing for companies that need one person to design the work and prove it worked.

The client was selling, but the conversion tracking had recorded zero purchases. Every decision about spend was being made blind, and the product feed was rejecting items faster than anyone was fixing them.

A catalog well over a thousand items deep, all inherited data, no budget increase available, and a live store that could not go dark while I worked on it.

Rebuilt the Merchant Center integration end to end. Corrected the catalog data, fixed the feed configuration, then repaired the conversion tracking so purchases were actually recorded against the clicks that caused them.

Nothing about the offer changed. The store had been invisible in Shopping and unmeasurable in Ads at the same time, so fixing the plumbing released demand that already existed.

Leads arrived with no source attached. The client could see totals and nothing underneath them, so every budget conversation was an argument about opinions.

A fixed ad budget that could not grow. Any improvement had to come from reallocation, not from spending more.

Built full-funnel tracking with conversion pixels and Google Tag Manager parameters. When the off-the-shelf tools could not carry click IDs into the lead forms, I wrote a WordPress plugin that captures UTM and click ID data from Google Ads directly into WPForms.

Once every purchase traced back to a specific click, the losing campaigns were obvious. Moving money toward what was already converting did the rest.

Six verticals and two affiliate entities were sharing one site that described the firm the way the firm described itself, rather than the way buyers looked for it.

Everything published had to clear FINRA review, so speculative claims and loose language were not available to me.

Rewrote most of the static content, rebuilt the metadata around real search demand, and cleaned up the interface so a visitor could find one vertical without reading about the other five. Then I paired search data with the firm's CRM to decide what to publish next.

The CRM knew which conversations turned into deals. The search data knew what people typed. Putting those two together turned content planning into a decision instead of a calendar.

A global rebrand needed to reach markets with no local web presence at all, and regional teams were producing their own materials with no shared rules.

Dozens of countries, multiple languages and time zones, and no appetite for a central bottleneck that would slow every region down.

Launched 42 international corporate websites and wrote the global brand standards behind them. Then I taught regional teams and the sales organization to work inside those standards, including social selling, so the system survived without me approving every asset.

Standards only hold when the people using them understand the reasoning. The training was the deliverable as much as the sites were.

Clients were starting to ask whether AI assistants were quoting them. Standard analytics could not answer it, because bot traffic gets filtered out before anyone looks at it.

No vendor sold the answer. If I wanted the data, I had to collect it myself.

Built a WordPress plugin that identifies and logs 44 AI crawlers across client sites, so we can see which systems visit, how often, and which pages they take. That sits alongside the AEO and GEO work: schema strategy, metadata, and content structured so a model quotes it accurately.

Discovery is moving inside AI answers. Being the source an assistant cites is a marketing problem, and it is one you can only work on if you can measure it.

Medical practices with several locations had business profiles scattered across unrelated accounts, inconsistent name and address data, and no way to advertise against local intent.

Healthcare advertising rules, plus profiles that had accumulated years of conflicting edits from people who no longer worked there.

Centralized every Google Business Profile into one master account, corrected name, address and phone consistency across directories, cleaned up citations, and launched verified Local Services Ads for each location.

Listing accuracy is unglamorous and decisive. A patient searching two miles away cannot choose a practice the map does not believe exists.