Apiary Foundry / operator-led growth system

Proof beats theater.

Apiary Foundry"s work is judged by systems built, data preserved, budgets moved, and execution timelines compressed. Here is what that looks like in practice.

Work with Apiary Foundry

operator reviewworkflow QA

Every page, workflow, and campaign has to produce a decision-ready signal.

The $50 middleware that made millions

Military.com — paid media turnaround

When Willie arrived at Military.com, the paid ad account was bleeding money so badly that he turned the ads off and nobody noticed for four months. The team had been spending tens of thousands of dollars and producing maybe three or four sales.

The problem was the conversion value problem: ad networks treated every conversion as the same, whether it was a $1 sale or a $100 sale. Machine learning algorithms need value signals to optimize for revenue, not just volume.

Willie found a $50-a-month Ukrainian startup called RedTrack. The middleware captured click IDs from the ad network, stored them in a database, matched them to actual transaction values, and sent the conversion values back to the platforms. Simple. Cheap. Effective.

Within three to four weeks, the account was producing roughly $1.40 for every $1.00 spent. They kept raising the budget until they maxed out search impression share. They made military recruitment ads profitable. They cross-sold leads across multiple branches — Army, Navy, Reserves, Marine Corps — because the system could now tell which traffic sources produced qualified demand.

The lesson: Tracking is growth infrastructure. A $50 middleware investment can become a multi-million-dollar competitive advantage when it preserves the data that decides budget allocation.

Tracking is growth infrastructure. Data retention changes the quality of every budget conversation.

43 consecutive profitable months

UpgradedPoints — travel credit card affiliate publisher

A travel credit card affiliate publisher had brilliant SEO but an underdeveloped paid media system. The founder was an SEO expert; the paid media infrastructure was not keeping up.

Willie worked with the founder and an internal developer team to build a RedTrack-equivalent in-house — middleware connecting click IDs to transactions and feeding conversion values back to Google Ads. Two months of build time.

The result: 43 consecutive profitable months without losing money. They maxed out Google Ads, moved to Meta, and scaled to over $1 million per month in profitable ad spend. At Meta they hit 1.1–1.2x ROAS — slightly lower than Google but still very profitable, running like clockwork.

That streak ended shortly after Willie left, when the company relocated the marketing team and the new administration lost money for five consecutive months. The 43-month proof point stands: the system worked when the operator understood how to build and maintain it.

Capabilities demonstrated

Click ID capture and conversion value preservation
Ad platform feedback loops (tROAS, value-based bidding)
Cross-channel scale (Google Ads → Meta)
Funnel economics and budget discipline
Downstream revenue signal routing

Building infrastructure from zero

The bank VA loan division

A bank hired Willie as a "mini CMO" to launch a new VA loan department. When he arrived, the reality did not match the job description:

No call center
No lead infrastructure to pass leads to sales
No CRM routing or stage architecture
Loan officers calling cold affiliate lists from cell phones
Massively overweighted sales staff with no qualified leads

Willie built the entire demand-to-close pipeline in about six months:

Call center
Lead piping and routing rules
Lead response workflows
Drip campaigns and nurture sequences
CRM integration and stage tracking

It took twenty different people and a lot of meetings. The infrastructure was solid. Then interest rates spiked overnight, the bank overleveraged itself on the VA loan bet, and the entire division — then the entire bank — shut down. The build was correct. The macro was not.

The lesson: You can build perfect infrastructure, but if the business model relies on cheap capital, even the best growth system cannot save it.

Two weeks vs. six months

First AI agent deployment on a marketing build

At a staffing agency division, Willie was given a chance to deploy AI agents on a marketing project for the first time. The constraints were tight: the parent company had an ADHD-octopus CEO with seven side projects and no coherent strategy.

Willie and his AI agents built:

50+ landing pages
150+ marketing emails
HubSpot CRM integration
N8N data tracking pipelines
Self-booking calendar system

All of it in two weeks. What took six months at the bank now took two weeks, and most of the delay was waiting for human reviewers to check the AI"s work or answer questions from developers who had built older infrastructure.

After the initial build, they pivoted multiple times, pulled calendars on and off, fixed broken tracking scripts written by human developers — all using AI agents. On the way out, Willie handed over a detailed runbook so the remaining team could operate the system.

The lesson: The right combination of deterministic systems, agentic execution, and human strategy review can compress build timelines by an order of magnitude without losing accountability.

What AF builds for clients

These proof points are not trophies. They are patterns. The same principles show up in every Apiary Foundry engagement:

Deterministic tracking — click IDs, UTMs, conversion values, and source data that survives from ad click to revenue event
Economic campaign structure — budgets moved by ROI signal, not vanity metrics
Connected systems — landing pages, forms, CRM, nurture, and reporting talking to each other
AI-assisted execution — agents and automation handling repeatable work, checked by review loops
Human strategy at the center — someone asking whether the work deserves money