Google Ads and landing pages: −38% cost per lead
A business that lives on leads from online advertising: about €400,000 on Google Ads since 2015, several brands, several websites. On one brand a new landing page cut the cost per lead from €10.7-14.2 to €7.4-8.1. Since September 2026 the daily analysis of the accounts is done by AI, which proposes changes with the numbers: they go live only after approval.
Before
Dozens of campaigns run by hand. Every day someone had to choose what to look at and what to ignore; changes were made in a hurry, with no single rule, and it was not always clear whether a drop in leads came from the campaigns or from a measurement error.
What I did
The landing page. On one brand I rebuilt the page the ads point to. With the same keywords and offer, the share of visitors who get in touch after the click rose from 5.3-6.9% to 8.6-9.3%, and the cost per lead fell from €10.7-14.2 (January-May 2026) to €7.4-8.1 (June-August 2026): about −38%. A lead here means a tap on phone, WhatsApp or email from the page.
Daily analysis with AI. I wrote the rules of the trade into a manual for AI agents: what to look at, what never to touch, how to propose a change. The agents read the accounts every day through Google’s official interface and propose with the numbers; they write only after an explicit yes. New campaigns start paused, budgets do not go up without confirmation.
Measurement. Every lead is tracked from the ad click to the email or phone call, with cookie consent handled properly.
After
- Cost per lead −38% on the brand with the new page.
- 42 landing pages generated from templates, each backed up and checked on a smartphone screen.
- Every decision is logged with the numbers behind it.
What it can do for your company
If you spend on Google Ads and do not know for sure what works: verified measurement, pages that convert, daily analysis and changes only with your yes. Fixed fee, never on commission.
Technical details, for those who want them
- Official Google Ads API, read and write, with pre-validation of every change; about 25 repeatable procedures for campaigns, budgets, schedules, negative keywords, extensions, conversions (11,439 lines of Python).
- Daily negative-keyword automation with safeguards, preview by default.
- Server-side tag manager, cookie consent, click-to-WhatsApp and click-to-call conversions, click ID carried through to the lead’s email.
- 195 code revisions between 6 and 28 September 2026; one session per account with a technical lock against wrong-account errors.
- Data: analysis of 14/9/2026 on the brand’s account, monthly values for January-May and June-August 2026.
The other systems I built.
Staff attendance with a WhatsApp message
Beforeattendance and messages in a WhatsApp group, hours counted by hand.
Afterthe employee writes “in”, the system records it, the administrator checks.
Websites, online shop and digital contracts
Beforeold sites hard to update, a messy catalogue, paper contracts.
Aftera custom platform for several brands, a clean catalogue, contracts signed on a tablet.
Do you have a similar problem?
Tell me in two lines: I will tell you whether the same approach applies to your company and what it would take to start.