Overview

Implementation Overview

Implementing Adobe Analytics means getting data from your website into Adobe's servers. There are multiple paths depending on your tech stack and requirements.

The implementation layer

Your Website
    ↓
Data Layer (JavaScript object)
    ↓
Web SDK or AppMeasurement (Adobe library)
    ↓
Adobe Edge Network
    ↓
Report Suite (database)

Modern vs legacy

Modern: Web SDK (recommended)

• Technology: JavaScript library, uses Adobe Edge Network
• Approach: Event-based, XDM schema
• Best for: New implementations, single-page apps, omnichannel
• Performance: Optimized for speed and efficiency

Legacy: AppMeasurement (still used)

• Technology: Older JavaScript library, direct to Adobe Analytics
• Approach: Variable-based (s.prop1, s.eVar1)
• Best for: Existing implementations with minimal changes required
• Status: Supported but not recommended for new projects

You should use Web SDK unless you have constraints. It's newer, faster, and better aligned with Adobe's platform strategy.

Implementation methods

Method 1: Web SDK + Adobe Tags (recommended)

You use Adobe's tag manager (formerly called DTM) to deploy Web SDK without coding.

Pros:

• No code changes needed on your site
• Rules are configured in UI
• Data layer abstraction
• Easiest to maintain

Cons:

• Tag manager adds slight overhead
• Dependency on tag manager availability

Method 2: Web SDK + direct implementation

You implement Web SDK directly in your site code.

Pros:

• Full control
• No tag manager overhead
• Can integrate with build tools (webpack, etc.)

Cons:

• Requires developer involvement for changes
• Harder to manage across multiple properties

Method 3: Server-side implementation

You send data to Adobe from your server (Node.js, Python, Java, etc.) instead of the browser.

Pros:

• Can use server-side data your browser doesn't have
• Harder for ad blockers to interfere
• More secure (API keys server-side)

Cons:

• More complex to implement
• Latency (data comes after page load)
• Only for specific use cases

The data layer

Before you send data to Adobe, structure it in a data layer — a JavaScript object that holds all the data.

window.adobeDataLayer = [
  {
    page: {
      name: "Homepage",
      type: "landing",
      siteSection: "home"
    },
    user: {
      id: "12345",
      segment: "member",
      isAuthenticated: true
    },
    event: "pageLoad"
  }
]

Web SDK or Tags read from this data layer and map it to Adobe's schema.

Why: Decoupling. Your site code doesn't need to know about analytics. You can change analytics logic without touching your site.

XDM vs variable-based

XDM schema (Web SDK)

Adobe uses a standardized schema (Experience Data Model) for all data:

{
  xdm: {
    web: { webPageDetails: { name: "Homepage" } },
    device: { type: "Mobile" },
    eventType: "web.webpagedetails.pageViews"
  }
}

Variable-based (AppMeasurement)

Analytics-specific variables:

s.pageName = "Homepage";
s.prop1 = "home";
s.eVar1 = "member";

Web SDK's approach is better because it's standardized across Adobe's ecosystem. But it requires learning new schema.

Environment strategy

You need three environments:

Key: Do NOT test in production. Ever.

Implementation checklist

Before launching:

• [ ] Data layer defined and stable
• [ ] Web SDK/AppMeasurement installed
• [ ] Props and eVars mapped to data layer
• [ ] Events configured
• [ ] Dev environment working
• [ ] QA completed in staging
• [ ] Monitoring/alerts set up
• [ ] Rollback plan documented
• [ ] Team trained

---

Next: Web SDK and the Edge Network