AI training & developmentCase study — built by one of our teams
Your marketing,
built in-house.
A team that sells on a large online marketplace could not see why its products rose or fell in the marketplace’s search. With one of our builders, the team built its own way to measure that and to test what moves it.
In progress (October 2026): the first listing test is running.
- 01
What the team needed · Starting point
See what moves search rank
The marketplace shares almost no search data. The team could not see where its products ranked, which search terms bring demand, or whether a listing change helped.
- 02
What the team built · 1:1 project work
Scans, a demand check, a test plan
Daily scans that record where its products rank for the search terms that matter. A demand check that ranks search terms by demand and by how many products compete. A written test plan: change one listing field on chosen products, keep a control group of similar products untouched, compare after five to seven days. A second track tests ads.
- 03
How the team runs it · Run by the team
Test, compare, keep what works
The scans run every day. A team member makes the listing changes the plan names. After five to seven days the team compares the changed products with the control group, logs the verdict and picks the next test. Our builder reviews the data with the team lead.
- Project data
- Baseline from the first 12 days of daily scans: on average 6.7 of the tracked products were in the top 120 search results. The best position was 32.
- Where our builder helped
- Our builder worked 1:1 with the team lead on the scans and the test loop. The team makes the listing changes and runs the tests.
You own it. No agency tax.
The full system
The whole funnel,
run by your team.
One way teams use the program: your own people build and run the whole system that brings in customers, paid and organic.
Illustrative: the examples below use invented data.
- 01
Start
Assess the whole funnel
A marketing assessment of the whole funnel, from first ad to sale, with an AI-visibility audit. Your team builds it and finds the weak parts.
- 02
The course, then your project
Build each piece with your team
Weak parts first, built by your people with us: paid ads with an AI creative-testing loop, the follow-up after the click, organic content, SEO and AI visibility.
- 03
Build and care
Measure, test, run
One weekly briefing measures every channel. Your team tests, keeps what works, automates it and runs it, on your own accounts and data.
The buying guide lifted sign-ups.
- What changed
- Newsletter sign-ups rose from 41 to 58 after the buying guide.
- LinkedIn reach fell 22%: two posts last week, not four.
- Ad tests
- Hypothesis held: the short video cut the cost per lead by 18%.
- Logged in the test ledger. Next week: two new opening lines.
- What to post
- LinkedIn Five short posts cut from the buying guide. Drafts ready.
- Newsletter The onboarding checklist, rewritten for small teams.
- What AI assistants said
Mentioned in 12 of 40 answers, up from 10. Assistant A now cites the buying guide.
Week 4: mentioned in 12 of 40 answers.
| Assistant | Week 1 | Week 2 | Week 3 | Week 4 |
|---|---|---|---|---|
| A | 2 of 10 | 3 of 10 | 3 of 10 | 5 of 10 |
| B | 1 of 10 | 1 of 10 | 2 of 10 | 3 of 10 |
| C | 0 of 10 | 1 of 10 | 1 of 10 | 1 of 10 |
| D | 3 of 10 | 3 of 10 | 4 of 10 | 3 of 10 |
| All | 6 of 40, 15% | 8 of 40, 20% | 10 of 40, 25% | 12 of 40, 30% |
You own it. No agency tax.