Building an autonomous growth engine on our own site
A live experiment — a system that finds growth opportunities, drafts changes and measures what happened. Running on vianapulse.com, with no results yet.
- Client
- Viana Pulse
- Industry
- Media and marketing
- Year
- 2026
- Status
- Live — in progress
The problem
Growth work is a loop — read the data, find the opportunity, make the change, wait, measure, decide what to do next. The loop is well understood. What makes it expensive is that every turn of it needs a person, and the waiting between turns means that person is never working on only this.
The approach
Rather than a content generator or another dashboard, build the loop itself: a central orchestrator that collects Search Console and analytics data, ranks opportunities, drafts a change, routes it for approval, publishes it, and records what happened at 7, 14 and 28 days. Autonomy increases in stages — the system proposes and a human approves before anything runs unattended.
Outcome
In progress. The site is live and the Search Console property is verified, which starts the baseline window. No growth has been measured and none is claimed here. Results will be added as the measurement windows close, each with its date and source. If the experiment shows that automation does not produce growth, that result will be published too.
Baseline — recorded 31 August 2026
Source: Google Search Console property for vianapulse.com, verified by the Owner
The site launched with no prior history, so the baseline is a genuine zero rather than a reconstructed figure. This is the one measurement that cannot be recreated later.
- Organic impressions at baseline0 — property verified at launch, no prior history
- Organic clicks at baseline0 — property verified at launch, no prior history
- Indexed pages at baselineSite launched 2026-08-31; indexing had not begun
Results
No results yet. Measurement windows are 7, 14, 28 days from the baseline date. Figures will be added as each window closes, with its date and source attached.
Why our own site
Testing a growth system on a client’s site would mean the first experiment carries someone else’s risk. Ours is new, we own the data, and the official APIs give us measurement we can trust. If the system does something wrong, it does it to us.
What is deliberately not automated
Redirects, canonical tags, noindex, deleting content and structural changes
all require human approval. These are the changes that are cheap to make and
expensive to undo, and they are exactly the ones an eager system would reach
for first.
The same applies to anything commercial. The system may draft how something is said; it may never change what is asserted to be true, or what is charged for it.
Why there are no numbers on this page
The baseline was recorded on the day the site launched. Real figures arrive when the first measurement windows close. Publishing projections now — even labelled as projections — would undermine the only thing this case study is for.
Evidence
Every action the system takes is logged with its hypothesis, the model and prompt version used, the expected result and the observed one, so any figure published here can be traced back to the run that produced it.