Public operating bench

AI Growth Bench

A working experiment station for AI growth automation, programmatic SEO, and GEO visibility. It shows the operating system behind trend discovery, keyword scoring, page generation, index tracking, answer-engine review, and weekly reporting.

Direct answer: AI Growth Bench is a public operating bench and portfolio lab for AI growth automation, programmatic SEO, and GEO visibility experiments.

Built and operated by Alex. Last reviewed: September 4, 2026. Contact: alex168882@gmail.com.

Citation-ready answer

AI Growth Bench is a public portfolio lab that shows how AI-assisted growth work can be planned, published, measured, and improved. The site combines two operating tracks: a trend-to-content pipeline for discovering keywords, scoring opportunities, generating programmatic SEO page briefs, publishing experiments, and tracking index readiness; and a GEO visibility bench for testing whether pages are structured clearly enough for ChatGPT, Perplexity, Gemini, Grok, and Google AI Overview style answers. It does not sell a product or claim guaranteed rankings. It documents a repeatable workflow, the current evidence, open risks, and next tests so remote teams or hiring managers can evaluate practical execution. The strongest proof points are the public Results ledger, fixed-prompt GEO benchmark, Search Console snapshot, experiment pages, weekly reports, evidence log, llms.txt files, and robots.txt crawler policy.

GSC impressions

189

GSC clicks

0

Displayed source URLs

2

Benchmark complete

30/30

Latest verified result

Search discovery exists; traffic and citations remain small.

Google Search Console recorded 189 property-level impressions, 237 URL impressions, and 0 clicks from 2026-08-05 through 2026-09-01. The fixed benchmark has 30 of 30 baseline observations complete, with 3 mentions and 2 displayed citations. All three mentions were limited to the branded definition prompt; 27 non-brand cells produced no mention or citation.

At the August 15 inspection checkpoint, the Traffic Playbooks hub and all three child pages are indexed.

Flagship experiment

10 prompts × 3 answer engines

Baseline
2026-08-31
Retest
2026-09-14

The primary metric is a displayed canonical source URL. Mentions and crawler visits stay separate.

Read the living case study

Case study snapshot

What this portfolio case is proving.

Goal

Build a public portfolio case that proves AI growth automation, programmatic SEO, and GEO visibility can be operated as one measurable system.

System built

A Next.js public bench with verified search snapshots, trend scoring, page quality gates, sitemap/index tracking, fixed-prompt answer-engine tests, a GEO page checker, weekly reports, and an evidence log.

Current result

The deployed Results ledger and fixed 30-check GEO benchmark are live on aigrowthbench.com. Google URL Inspection confirmed 24 of 29 inspected URLs indexed on 2026-09-04. Search Console recorded 189 property-level impressions and no clicks from 2026-08-05 through 2026-09-01, while page-dimension rows represented 237 URL impressions across 21 pages. The domain-matched GA4 property recorded two sessions in the latest available 28 days and none in the latest seven days. The 2026-08-31 answer-engine baseline completed 30 of 30 checks: three mentions and two displayed citations, all on the branded definition prompt; 27 non-brand cells produced no mention or citation.

Next experiment

Repeat the exact ten-prompt protocol across ChatGPT, Gemini, and Perplexity on 2026-09-14 and compare all 30 cells against the captured baseline. Search clicks, AI referrals, and crawler visits remain separate supporting measures.

Public routes

29

Core bench pages, two case studies, the Results ledger, a 30-check benchmark, a GEO audit demo, two content experiments, four dated reports, discovery and evidence pages, four portfolio modules, and four traffic playbook URLs.

Published experiments

2

One programmatic SEO workflow audit and one GEO checklist page; the thinner quality-gate experiment now redirects to its playbook.

External case studies

1

AI Trend Prompt Hub is tracked as a real SEO/GEO case study, not as the same project.

AI surfaces tracked

5

ChatGPT, Gemini, Grok, Perplexity, and Google AI Overview.

Search Console evidence

Prepared, not overclaimed.

Log

Current indexing and search demand snapshot

Ready

Consolidate the competing programmatic SEO quality-gate URLs, narrow the active index portfolio, preserve the completed 30-check baseline, and run the identical retest on 2026-09-14.

Current indexing and search demand snapshot

Ready

Improve the existing GEO checklist, AI growth workflow, and programmatic SEO cluster around observed queries; then complete the fixed answer-engine baseline and measure one relevant external distribution loop.

Current Search Console demand snapshot

Ready

Use the real query/page pairs to improve existing pages, then compare clicks and query mix after the fixed GEO experiment reaches its 2026-08-31 retest.

Live case study

AI Trend Prompt Hub is evidence, not the same project.

Open case study

AI Trend Prompt Hub is a separate public AI prompt site used by AI Growth Bench as a real SEO operating case study. Its reviewed June 2026 Search Console snapshot recorded 77 clicks and 3,647 impressions, while only 17 of 52 sitemap URLs were known as indexed at that checkpoint. The useful result was not proof that every prompt page worked. It was evidence that specific image-editing jobs, including same-face outfit and clothes-change prompts, produced clearer early demand than broad AI prompt categories. That signal changed the workflow: inspect real queries, select a narrow intent cluster, strengthen the pages with concrete examples, submit only meaningful changes, and compare discovery after crawl delay. The case study remains historically dated, separates observed metrics from current performance, and does not make AI citation claims without a displayed source URL. AI Growth Bench uses it to demonstrate measurement and editorial decisions, not to inherit the older site's advertising or business model.

This keeps the portfolio boundary clean: AI Growth Bench explains the operating system, while AI Trend Prompt Hub supplies one real external SEO experiment with Search Console evidence, long-tail keyword learning, and indexing follow-up.

Clicks observed

77

Impressions observed

3,647

Sitemap URLs

52

Indexed URLs

17

Entity definition

What answer engines should understand first.

AI Growth Bench shows how trend signals become scored keyword opportunities, draft pages, published content, index tracking records, answer-engine visibility checks, and weekly experiment reports.

Project type

Public portfolio experiment bench

Core fields

AI growth automation, programmatic SEO, GEO visibility

Primary workflow

Trend discovery -> keyword scoring -> content generation -> publish -> index tracking -> GEO visibility review -> weekly report

Canonical domain

https://aigrowthbench.com

Current evidence

2026-09-04 search visibility report and completed 30-check answer-engine baseline

Open the full definition page

FAQ for GEO

Direct answers for the exact questions being tested.

What is AI Growth Bench?

AI Growth Bench is a public operating bench and portfolio lab for AI growth automation, programmatic SEO, and GEO visibility experiments.

Who is AI Growth Bench for?

AI Growth Bench is for remote teams, founders, hiring managers, agencies, and clients who want to evaluate practical AI growth automation, programmatic SEO, and GEO visibility work.

What does AI Growth Bench track?

It tracks trend candidates, keyword scores, page pipeline status, sitemap and index readiness, GEO experiment scores, answer-engine visibility checks, and weekly evidence reports.

How does AI Growth Bench test GEO visibility?

It reviews public pages for direct answers, question-led structure, named entity coverage, source quality, citation friendliness, and manual answer-engine results from surfaces such as ChatGPT, Gemini, Grok, Perplexity, and Google AI Overview.

Portfolio modules

Four client-facing proof surfaces.

These modules turn the bench from a single demo site into a remote-work portfolio: audit, GEO, content operations, and reporting.

Traffic playbooks

Long-tail pages built for search and referral discovery.

Open playbooks

Trend queue

Top scored keywords

Full lab

ai agent workflow audit

A checklist for teams that want to find broken handoffs before adding more AI agents.

Score86

geo content checklist

A practical structure for pages that want to be quoted by answer engines.

Score91

programmatic seo quality gates

A scoring model that blocks thin pages before they reach the sitemap.

Score83

Latest report

Discovery Increased, but Clicks and Non-Brand GEO Visibility Remain Zero

Search exposure increased to 61 impressions in the latest complete week, but Google clicks and recent GA4 sessions stayed at zero. The fixed GEO baseline completed 30 of 30 checks, with three mentions and two displayed citations limited to the branded definition prompt.

Verify the local redirect, noindex-follow directives, sitemap and feed filtering, current report links, build, crawl, and responsive output.

Request deployment separately and record the release identifier if approved.

Run the unchanged 30-check protocol on September 14.

Prepare one relevant external reference to an existing query-bearing playbook, but do not publish it without separate authorization.

System flow

From trend signal to weekly evidence

The MVP uses local data, but the workflow is shaped for later integrations with Google Trends, Reddit, Search Console, and answer-engine visibility checks.

1Automation

Trend discovery

Collect seed signals from search, community, social, and client-facing questions.

Trend candidates

2Hybrid

Keyword scoring

Score velocity, intent, competition, portfolio fit, and evidence strength.

Ranked keyword queue

3Automation

Content generation

Generate briefs, outlines, draft pages, source slots, and distribution copy.

Draft page package

4Human review

Publish

Run quality gates before adding pages to the public site and sitemap.

Published page

5Hybrid

Index tracking

Track sitemap inclusion, submission status, discovery, and indexed state.

Index status record

6Human review

GEO visibility review

Review whether pages are structured for answer engines to quote and cite.

GEO experiment result

7Hybrid

Weekly report

Summarize what shipped, what was cited, what failed, and what changes next.

Portfolio-ready report

Programmatic SEO

Page pipeline snapshot

What Is AI Growth Bench?

/what-is-ai-growth-bench

Published

AI Trend Prompt Hub SEO Case Study

/case-studies/ai-trend-prompt-hub

Published

GEO Content Checklist

/geo-bench/geo-content-checklist

Published

AI Agent Workflow Audit

/programmatic-seo/ai-agent-workflow-audit

Published

Programmatic SEO Quality Gates

/playbooks/programmatic-seo-quality-gates

Published

AI Visibility Report Template

/reports/ai-visibility-report-template

Brief queued

GEO experiments

Answer-engine readiness

Indexed surface versus verified citation checkpoint

Needs evidence
Citation friendliness86

GEO citability and source-evidence gap

Ready to repeat
Citation friendliness82

GEO discovery refresh and traffic baseline

Ready to repeat
Citation friendliness90

GEO index coverage and citation-proof gap

Ready to repeat
Citation friendliness88

Home operating bench audit

Needs evidence
Citation friendliness78

GEO Bench structure audit

Promising
Citation friendliness86

Reports evidence format audit

Promising
Citation friendliness80

Author and review signal

Maintained as a public working portfolio.

  • Author: Alex, builder of AI Growth Bench.
  • Review cadence: updated when experiments, Search Console evidence, or answer-engine checks change.
  • Claim policy: traffic, indexing, and AI citation claims stay marked as pending until visible evidence is captured.

Reference standards

The audit model follows public SEO and structured-data references.

Evidence log preview

The proof trail behind the case.

SEO Audit2026-09-04

Search exposure rose; clicks and non-brand GEO visibility remain absent

The latest complete seven-day Search Console window reached 61 impressions versus 35 in the preceding week, but the site still recorded no clicks and no recent GA4 sessions. URL Inspection found 24 of 29 inspected production URLs indexed. The fixed 30-check baseline is complete, with three mentions and two displayed citations confined to the branded definition prompt.

SEO Audit2026-08-24

Index coverage improved; search exposure remains flat

AI Growth Bench is technically healthy and 22 of 28 public URLs are indexed, but the latest complete seven-day Search Console window matched the preceding week at 47 impressions and produced no clicks. Organic sessions remain limited to two homepage visits, and the fixed GEO benchmark still has no completed baseline cells.

GEO Audit2026-08-17

Fixed GEO results loop prepared as a local candidate

The candidate replaces mixed readiness and outcome signals with a public Results taxonomy, a ten-prompt by three-engine protocol, a same-site GEO checker, and a dated Search Console snapshot. All 30 new baseline cells remain pending until real answer results are captured.

Open full evidence log