---
title: Reading Weekly Trends
description: Interpret weekly changes in observed AI requests without treating them as citation or referral proof.
audience: [human, agent]
type: guide
version: v1
stability: stable
lastVerified: "2026-08-01"
prerequisites: [/docs/evidence-model/]
related: [/docs/concepts/ai-traffic-types/, /docs/api/analytics-endpoints/]
artifacts: []
---

## Overview

The weekly trend view aggregates classified requests so you can see changes without as much day-to-day noise. Always confirm the selected domain, date range, and request class before comparing periods.

## What to look for

### AI request share

The **AI %** column is the share of observed requests classified as AI-related within the selected data. A change in crawler share proves a change in observed requests. It does not by itself prove discovery, citation growth, or referral growth. Compare crawler visits with sampled citation and detectable referral evidence before deciding what changed.

### Platform shifts

The per-platform breakdown can reveal:

- **New observed identities** — a classified platform appears in the selected window.
- **Request spikes** — a platform makes substantially more observed requests.
- **Request declines** — observed requests fall, which may justify checking crawler access, deployment health, or classification changes.

These patterns are diagnostic leads. They do not establish why the platform changed its behavior or how generated answers used the pages.

## Using the API

```bash
curl "https://api.seellm.link/api/analytics/weekly-trend?range=90d&domain=example.com" \
  -H "Authorization: Bearer sk_live_..."
```

## Using MCP

Ask your AI assistant:

> "Show me the weekly observed AI request trend for example.com over the last 90 days"

The `get_weekly_trend` tool returns the weekly summary and per-platform breakdown.

## Acting on trends

| Observation | Check next |
|---|---|
| Crawler requests increase | Which important pages were requested, then compare sampled citations |
| One platform dominates | Classification evidence and page distribution for that platform |
| Requests decline | `robots.txt`, adapter health, and deployment changes |
| AI referrals increase | Landing pages, conversions, and whether referrer coverage changed |

Use the [Evidence Model](/docs/evidence-model/) when communicating what each trend does and does not prove.