Home Use Cases Work Insights About Contact
Back to Our Work

How a Growth-Stage HR Services Company Built a Content Loop Where Lead Results Set Next Week's Topics

Marketing1
Open the walkthrough

Five-Step AI Workflow with Human Oversight

Each step shows the workspace, what the AI completed, and what still needs human review.

Steps 1 and 3 need a person. Steps 1 and 3 are waiting for approval right now. Click either to open the full walkthrough.

Estimates are directional and based on stated assumptions. All names, organizations, and identifying details have been anonymized in accordance with our confidentiality agreements.

The Transformation

How a Growth-Stage HR Services Company Built a Content Loop Where Lead Results Set Next Week's Topics

Before
A person used to write every article, at about one a week.
A person used to load each article into the site and each channel by hand.
A person used to pull the numbers into a spreadsheet at month end.
After
the search gaps found, and the week's briefs waiting for your pick
the brief written into a full draft, in the company's voice
every check run, and the flagged sentences laid in front of you
the calendar filled, and every channel served at its slot
every live article measured, and the lessons handed back to the topic queue
Executive Summary

Turns the gap between what buyers search for and what the site answers into a steady stream of published articles that sound like the company, and never lets an unsourced number or an off-voice sentence carry the company's name. Five stations, run as a loop: what station 5 learns about which topics earn leads changes what station 1 puts in front of you next week.

What Was Broken

A person used to write every article, at about one a week
A person used to load each article into the site and each channel by hand
The real cost
A person used to pull the numbers into a spreadsheet at month end.

What We Built

Five stations and 10 subagents. Each subagent carries its own tasks and its own refusal.

A4
Keyword Watch
Pull the terms this site already ranks for, and where each one sits this week
A1
Brief Drafting
Draft a one-page brief per gap: the angle, the terms, the questions the page must answer
A1
Section Writing
Draft the article from its brief, section by section, in the voice the samples carry
A3
Voice Check
Compare every draft against the approved voice samples, sentence by sentence
A3
Page Check
Check the title, headings, meta description and links against the on-page checklist
A3
Claims Check
Verify every cited number against the source it cites
A1
Variant Drafting
Draft two title and meta-description variants per article for the live test
A4
Release Scheduling
Slot each approved article into the calendar, spaced so topics do not collide
A4
Traffic Attribution
Track rankings, traffic and time on page for every live article, every week
A3
Test Reading
Score every live title test against its traffic split

How It Runs

1

Topic Research

A person answers here

Agents track every term this company and its competitors rank for, find the gaps worth filling, and draft a one-page brief per gap before a person picks any of them. A person picks which briefs get written this week. That choice is strategy, and it stays a person's.

2

Article Drafting

Nobody is asked anything here

Agents write each commissioned brief into a full draft in the company's voice, with the target terms worked in where they read naturally and a source beside every number. Nobody writes or edits here. A person reads the finished draft at the review desk, not the half-finished one.

3

Editorial Review

A person answers here

Agents run every on-page check, trace every claim to the source it cites, and lay the draft in front of a person with only the flagged sentences left to read. A person reads the flagged sentences and approves the draft or sends it back. Every article crosses this desk before it carries the company's name.

4

Publishing

Nobody is asked anything here

Agents slot each approved article into the calendar, draft the two title variants its live test will run, and push it to the site, the newsletter and LinkedIn at its slot. Nobody schedules or posts here. The approval at station 3 is the last human touch before the article is live.

5

Performance Tracking

Nobody is asked anything here

Agents track every live article's rankings, traffic and leads, call each title test only when the numbers can carry the call, and route what worked back into the topic queue. Nobody assembles a report here. The monthly readout arrives written, and the findings are already in next week's topic ranking when a person opens it.

Where a Person Decides

Step 1, Topic Research. Pick the briefs that get written this week. Kill any topic that does not serve the pipeline, and say why, so next week's ranking carries it. no article nobody chose.
Step 3, Editorial Review. Read the two flagged sentences and settle each one. Approve the draft, or send it back with the flag that stopped it. the company's name is on every claim.

Operating Model

This changes how work flows through the team.

Role
Responsibility
Topic Research owner
Pick the briefs that get written this week. Kill any topic that does not serve the pipeline, and say why, so next week's ranking carries it. no article nobody chose.
Editorial Review owner
Read the two flagged sentences and settle each one. Approve the draft, or send it back with the flag that stopped it. the company's name is on every claim.

What Transfers, What Must Be True

What transfers
A person is in the loop where the strategy is chosen and where the company's name goes on a claim. Two of the five stations refuse to proceed on their own, Topic Research and Editorial Review, and those are the two a person carries. Everywhere else the agents run at volume and reach you only when they cannot settle something.
Topic Research stops for a person, and no article nobody chose.
Editorial Review stops for a person, and the company's name is on every claim.
Every subagent says what it will not do. 10 of them do.
What must be true in your environment
The agent can read the systems your records already live in. This one reads 16.
Somebody owns Topic Research and has time for it.
Somebody owns Editorial Review and has time for it.

Failure Modes

What breaks this pattern:

✗ Two pages answer one topic

The site already answers the topic, and nobody points at the existing page. The new article and the old page compete for the same searches, and both rank worse.

✗ The gap gets written anyway

No source carries the claim, so the draft fills the gap with something that sounds right. A reader checks it, and the article loses their trust.

✗ The calendar overrides the approval

The publish date arrives and the article goes out with no approval record. The first person to catch the error is a customer, not a reviewer.

✗ Thin traffic picks a winner

A few dozen visits look like a result, so the test names a leader. The team ships the wrong variant everywhere and never finds out.

Directional Outcomes

What the agent counts, and the station that counts it.

These counts are the tallies from one monitored run of the agents. They are not monthly or annual totals.

Drafts written
Counted at Article Drafting
41
Numbers written without a source
Counted at Article Drafting
0
Claims left marked for a person to source
Counted at Article Drafting
12
Articles and channel posts sent out
Counted at Publishing
114
Published without an approval record
Counted at Publishing
0
Held and sent back
Counted at Publishing
1
Articles tracked
Counted at Performance Tracking
214
Tests called on thin traffic
Counted at Performance Tracking
0
Drops handed to a person
Counted at Performance Tracking
1
Our measurement policy: We do not publish precise ROI without baseline methodology. Every figure above carries its basis.

What Runs Where

Every step names the subagent that does the work, the record it writes, the thing that raises a question for a person, and what it is allowed to touch. This is drawn from the source, not from a diagram somebody kept in sync by hand.

1Topic Research
GATE
subagenttopic-research
writespipeline/briefs/<topic-id>.json
raisescommission-briefs
may touchpipeline/**, rankings/** read-only
2Article Drafting
subagentdraft-article
writespipeline/drafts/<article-id>.md
raisesclaim-without-source
may touchpipeline/**, voice/** read-only
3Editorial Review
GATE
subagentreview-draft
writespipeline/reviewed/<article-id>.json
raisesawait-approval
may touchpipeline/**, sources/** read-only
4Publishing
subagentpublish-article
writespipeline/published/<article-id>.json
raisesapproval-record-missing
may touchpipeline/**, calendar/**
5Performance Tracking
subagenttrack-performance
writesreports/monthly/<period>.json
raisesunexplained-drop
may touchreports/**, pipeline/** read-only

Stack

Every system this agent reads or writes.

System
Read at
Stations
Google Analytics
Performance Tracking
1 of 5
Google Search Console
Topic Research
2 of 5
the CRM where the leads land
Performance Tracking
1 of 5
the LinkedIn company page
Publishing
1 of 5
the WordPress drafts folder
Article Drafting
2 of 5
the WordPress site
Publishing
1 of 5
the approved voice samples
Article Drafting
1 of 5
the brief library
Article Drafting
1 of 5
the cited sources
Editorial Review
1 of 5
the competitor rank tracker
Topic Research
1 of 5
the email newsletter
Publishing
1 of 5
the on-page checklist
Editorial Review
1 of 5
the publishing calendar
Publishing
1 of 5
the title-test results
Performance Tracking
1 of 5
the top-ranking pages for the term
Editorial Review
1 of 5
the topic queue
Topic Research
1 of 5
Next Step

Want to see if this pattern fits your marketing?

No build commitment·Real samples, not a demo·Estimate in writing