---
title: Growth Engineer
category: product
entity_type: agent
price: $79
canonical: https://forgehouse.ai/agents/growth-engineer/
lang: en
hreflang_alt: https://forgehouse.ai/tr/ajanlar/growth-engineer/
last_updated: 2026-06-20
---

# Growth Engineer

> Paid ads, tracking and retention specialist

A growth specialist that thinks in ROAS, CAC and LTV. It structures Google and Meta ad accounts, hunts wasted spend, audits conversion tracking, and designs funnels and retention loops. New campaigns are born paused behind an approval gate, conversion counts get cross-read from two sources before any budget call, and the search-terms report is the first stop on every inherited account, waste is found before it is funded.

## Use cases
- Auditing a Google or Meta Ads account for wasted spend
- Structuring a new campaign with the right keywords and negatives
- Verifying conversion tracking across Ads and GA4
- Designing a funnel or CRO experiment
- Building a retention / win-back sequence strategy

## Benefits
- Stop burning budget on searches that never convert
- Trust your conversion numbers across every channel
- Lower cost per acquisition with cleaner account structure
- Decisions driven by ROAS, not vanity clicks

## What’s included
- Account-structure design for Google and Meta with phrase-match discipline
- Wasted-spend hunting: search-terms report, n-gram clustering, negative lists
- Conversion-tracking audits: platform count cross-checked against GA4 every time
- New campaigns born PAUSED behind an approval gate: no silent spend
- Funnel and retention reads: CAC/LTV/ROAS framing on every recommendation
- Budget reallocation proposals backed by two data sources, never one

## Who it’s for
Advertisers who want every ad dollar measured and defensible.

## How it runs
Most ad accounts measure conversions wrong, so this agent lists conversion actions before holding any CPA opinion. Ads and GA4 numbers sit side by side in every report, and live campaigns are never touched without approval.
1. Invokes the MCP data toolkit before the first API call to load client account IDs, customer ID formats and known data traps, then reads the agency Google Ads or Meta standard with its field lessons, including the systemic finding that most accounts have broken or inflated conversion tracking, so conversion actions are listed before any CPA opinion.
2. Pulls real account data for the task: campaign performance, the search terms report, the wasted spend report and keyword quality scores for an audit; GA4 funnel, landing pages and key events for tracking and CVR work.
3. Runs the mandatory conversion cross-check: Ads conversions next to GA4 conversions in every report. A 0 to 15 percent gap is normal attribution difference, 15 to 30 is suspicious and triggers a tracking audit, above 30 means the tag is broken and nothing gets enabled.
4. Hunts wasted spend by splitting search terms into 2-gram and 3-gram clusters that spend without converting, checks the quality score triad (keyword, ad headline, landing page carrying the same intent) and rejects Google's self-serving suggestions like broad match and auto-applied recommendations by default.
5. For a new campaign, runs the three-gate flow: an autonomous setup brief with structure, phrase-only keywords, starter negatives and a daily budget proposal; written CEO approval; then creation in PAUSED state, with going live requiring a separate explicit command. Live campaigns are never touched without approval.
6. Closes through a 5-point gate: every metric claim backed by a real API pull with a verification tag, the Ads-to-GA4 table present, each recommendation framed as problem plus evidence plus impact plus action, and client-facing output free of jargon and speculative currency figures.

## FAQ
### I inherited a messy Google Ads account from another agency, where does this agent start?
The search-terms report, every time. On an inherited account it hunts wasted spend first: n-gram clustering of search terms, negative-keyword lists and a structure review with phrase-match discipline. Budget recommendations only come after the waste map exists, so you stop funding searches that never convert.

### How can I trust its conversion numbers more than my own dashboard?
Because it refuses single-source stories. Platform conversion counts are cross-checked against GA4 before any budget call; a large gap between the two triggers a tracking audit instead of a spend decision. Every recommendation is framed in CAC, LTV and ROAS rather than clicks.

### Will it launch campaigns and start spending on its own?
No. New campaigns are born paused behind an approval gate by design, nothing goes live or spends without your explicit go. It also does not write your landing pages or long ad copy; it defines the structure, targeting and measurement around them.

## Price
$79, one-time, no subscription. VAT included.

Related guide: [AI Google Ads and Meta Ads management](https://forgehouse.ai/guides/ai-google-ads-management/)