Docs · v1

From signup to your first receipt in under 10 minutes.

Everything you need to install Aelo, capture your first scans, and start acting on the answers. Prefer to see the formulas behind the numbers? Read the methodology.

Quickstart Install the pixel Event API Scan API Core concepts FAQ

Section 01

Quickstart

1. Create your workspace

Sign up (no card). Auto-onboarding provisions your first workspace with sensible defaults. Add your website and top 3 competitors in the first minute, the analyzer uses them to categorize citations correctly.

2. Add target prompts

LLM Tracker → Add prompts. Aelo seeds 30+ high-intent queries for your category. Edit, star, or add your own. Every scheduled scan runs against every starred prompt.

3. Run your first scan

Click Run scan. Aelo hits Gemini live (Radar tier) or all four providers in parallel (Command). Analyzer extracts mention state, position, sentiment, competitors, and citations. Zero fabrication, a failed provider surfaces a real error, never a mock.

4. Log your first intervention

Ship something, a landing page, a Reddit reply, a schema block. Log it in /dashboard/interventions with the target prompt(s) it should move. Aelo snapshots a baseline at that moment. Hit Measure a week later to see the receipt.

Section 02

Install the pixel

The Aelo pixel is a single async script that measures which AI referrers drive real visits, ChatGPT.com, Perplexity, Gemini, Claude, and Google AI Overview all set identifiable referrers, and the pixel classifies them for you.

Basic install

<script
  id="aeo-pixel"
  src="https://aelohq.com/aelo-pixel.js"
  data-workspace-id="YOUR_WORKSPACE_ID"
  async
></script>

Paste before </head>. Grab the workspace-personalized snippet from Settings → Install , the workspace ID is filled in for you and the tab shows a live verification badge that flips to VERIFIED the moment the first pageview arrives.

Framework-specific placement

Next.js (App Router)

Add to app/layout.tsx inside <head>, or use next/script with strategy="afterInteractive".

Next.js (Pages Router)

Add to pages/_document.tsx inside <Head>.

React / Vite / CRA

Add to public/index.html.

Webflow / Framer

Site Settings → Custom Code → Head Code.

Shopify

Themes → Edit code → theme.liquid.

WordPress

Theme File Editor → header.php, or use a plugin like Insert Headers and Footers.

What the pixel sends

POST https://aelohq.com/api/analytics/track
Content-Type: application/json

{
  "workspaceId": "wsp_XYZ",
  "eventType":   "pageview",
  "url":         "https://your-site.com/pricing",
  "referrer":    "https://chatgpt.com/",
  "aiSource":    "chatgpt",         // null when not from an AI referrer
  "userAgent":   "Mozilla/5.0 …",
  "timestamp":   "2026-07-05T14:23:15Z"
}

No cookies. No fingerprinting. No personal data. Just enough to know which LLM sent the visitor and what page they landed on.

Section 03

Custom events

The pixel exposes window.aelo.track() so you can log key conversion events with the AI-source attribution attached. Signups, purchases, form submits, whatever you want to close the attribution loop on.

Fire a custom event

window.aelo?.track("signup", {
  plan: "command",
  ltv:  4900,          // your own custom fields, freeform
});

Shape of the payload sent to Aelo

{
  "workspaceId": "wsp_XYZ",
  "eventType":   "signup",           // your event name
  "aiSource":    "chatgpt",          // hydrated from initial pageview
  "metadata":    { "plan": "command", "ltv": 4900 },
  "timestamp":   "2026-07-05T14:23:47Z"
}

Events attributed to an AI referrer show up in Dashboard → Attribution alongside pageviews.

Section 04

Scan API

Programmatic access to run scans and pull receipts. Get an API key from Settings → API Keys.

Run a scan

POST https://aelohq.com/api/llm/scans
Authorization: Bearer aelo_sk_...

{
  "prompt":     "Best CRM for Indian SMBs in 2026",
  "brandName":  "Zoho",
  "brandDomain":"zoho.com",
  "platforms":  ["gemini", "chatgpt"],
  "competitors":["Freshworks", "HubSpot", "Salesforce"]
}

List recent scans

GET https://aelohq.com/api/llm/scans?limit=50&platform=gemini
Authorization: Bearer aelo_sk_...

All scan endpoints are rate-limited (20 requests / hour / API key on Command). Failed scans return a real error surface (429, 5xx, etc.), never a mock success. Full response schema is documented in the API repo.

Section 05

Core concepts

Mention rate

Percentage of tested prompts in which the AI named your brand. Deterministic string match; no LLM-as-judge.

Average position

When named, how early in the AI's list. Position 1.0 = first mentioned. Scans where brand wasn't named are excluded from the average (not counted as ∞).

Health score / 100

Composite of mention rate (weighted 70%) and position boost (30%). 0 = never mentioned; 100 = named first every time. Formula on /methodology.

Share of Voice

Your mentions as a fraction of all named brands in your competitor set. Requires competitor list configured; otherwise displays ', '.

Receipt

The raw scan behind any metric, prompt sent, LLM response received, timestamp. Every derived number in Aelo is one click away from its receipt.

Intervention

Something you did to improve visibility (a landing page, a Reddit reply, a schema block). Aelo snapshots a baseline when logged; Measure a week later to see the lift.

Read all the formulas

Section 06

FAQ

How is Aelo different from ChatGPT SEO tools?

Most tools ship a black-box 'AI Score' averaged across platforms. Aelo shows per-platform per-prompt receipts you can reproduce yourself in 30 seconds. See /methodology for every formula.

Does the pixel slow down my site?

The script is ~3 KB gzipped, loaded async, and fires a single beacon per pageview. No render-blocking, no cookies. See the Event API below for the exact wire format.

Which LLMs does Aelo scan?

Gemini and Google AI Overview on Radar. ChatGPT, Gemini, Claude, and Perplexity on Command. Same prompt runs on every enabled platform in parallel, then each response is analyzed independently, no cross-platform averaging unless you ask for it.

What happens when an LLM API is down?

The failed scan is logged with the actual error, tagged failed, and excluded from all metrics. It does not become a 'zero' or 'not mentioned' data point. Failures are visible in the receipt drawer.

Can I export my scan data?

Yes, every workspace can export every scan as CSV or JSON from Settings → Data. Nothing about a metric is proprietary; take the receipts with you if you leave.

How much does the pixel cost me on the LLM side?

Nothing, the pixel captures pageviews on your site. LLM scans are billed against Aelo's plan. Your visitors never trigger an LLM call.

Something not covered?

Email us at docs@aelohq.com. We reply within a day, and every question becomes a doc entry.

Book a walkthrough