Robust Branding experiment log
We run every service on our own group before we sell it. The log records the brief, the spend, the result and the decision. Spend and clicks update daily while a test runs, and the result and decision publish on its readout date.

The contact sheet for our current tests
Numbers update daily. The next readout is today.
Home and field services
45 threads started, 71 emails sentProfessional firms
34 threads started, 48 emails sentEnergy
37 threads started, 67 emails sentData center builders
5 threads started, 13 emails sentManufacturers and distributors
41 threads started, 55 emails sentAccessibility
EUR 55.05 spent, 5 clicksAssistant Answers
EUR 35.13 spent, 3 clicksDrupal 10 End of Life
EUR 11.90 spent, 2 clicksLocal Presence
EUR 23.38 spent, 3 clicksSite Operations
EUR 31.29 spent, 4 clicksTechnical Search
EUR 0.00 spent, 0 clicksEvery experiment we have run
Running tests first, then the next in the queue, then each finished test with its result and decision.
Running now
Outbound to the same five industries
A five-email sequence per account sent from dedicated domains. A person reads every reply and a daily index counts them. We run it beside RB-001 to compare search and outbound per industry.
Six monthly plans on Google Search
Six plan pages with a published monthly price, each on its own capped budget. The keywords name the service a buyer hires, and a visitor sees the price above the enquiry form.
Three articles take turns on one search campaign
One search campaign keeps its keywords and its EUR 10 daily budget and sends value creation consulting searches to an article on Async Advisor, then Value Creation Partner, then Growth Shuttle, two weeks each. One ad serves at a time so the three pages never bid against each other.
Held. No ad is serving, so the window counts days without spend.
Completed
Five industry positions on Google Search
Five landing pages for home and field services, professional firms, manufacturers and distributors, energy, and data center builders. Each arm had EUR 140 and the same maximum bid in the US, UK and Canada. We paused the test on its first evening to add proof to the pages and restarted three arms on Nov 10.
Stopped on Nov 11. The searches we paid for came from people looking for a site builder, such as "plumber website", and not from companies that buy our work.
Result. EUR 58.69 of EUR 700 spent. 192 impressions, 10 clicks and no leads counted by the ad account. Home and field services took 7 of the 10 clicks.
Decision. We stopped buying search for these five positions. Outbound to the same industries continues as RB-002.
Experiment RB-001 by arm
Experiment RB-001 compared five industry positions on Google Search
US, UK and Canada. Nov 1 to Nov 11, 2026. Stopped early. Readout Nov 15.| Arm | Impressions | Clicks | CTR | CPC |
|---|---|---|---|---|
| Home and field services | 23 | 7 | 30.43% | EUR 5.89 |
| Professional firms | 35 | 1 | 2.86% | EUR 5.81 |
| Manufacturers and distributors | 131 | 1 | 0.76% | EUR 5.80 |
| Energy companies | 0 | 0 | ||
| Data center builders | 3 | 1 | 33.33% | EUR 5.88 |
| All five arms | 192 | 10 |
5 of 5 GTM5 campaigns read from Google Ads account 1955929444, 2026-11-01 to 2026-11-11; a campaign with no impressions shows zero. Updated 2026-11-15 14:25 UTC. The result and the decision are in the log above.
Client experiments in the log
We publish client results only with permission and anonymise them by default.