All case studies
Enterprise & Personalization

Unlocking Server-to-Server Personalization for a Fortune 500 Enterprise

A Fortune 500 tax preparation firm launched tamper-proof server-side pricing experiments — with durable identity, full Adobe Analytics reporting, and zero new infrastructure.

Days
From initiation to live experiments
~2 yr
Identity cookie lifespan vs. 7-day ITP
Zero
Infrastructure provisioned by customer
Zero
Customer impact from vendor SDK migration

The challenge

The firm's digital team needed to A/B test product pricing on high-traffic landing pages using Adobe Target. Two fundamental constraints made client-side delivery unacceptable.

First, security: any personalization delivered via JavaScript is visible and modifiable by the end user. Browser developer tools allow inspection and alteration of DOM content, JavaScript variables, and network responses. For a pricing experiment, this is a disqualifying risk — test values cannot be exposed in the browser.

Second, identity: modern browser privacy policies (Safari ITP, Firefox ETP) limit client-side cookies to a maximum lifespan of 7 days. A visitor who returns after 10 days is treated as new, re-bucketed into a different test group, and potentially shown a different price. This breaks both test validity and customer trust.

Building server-side personalization internally would require provisioning infrastructure, engineering an integration with Adobe Target's API, implementing durable identity, and coordinating across development, IT, and security teams — months of effort before a single experiment could run.

The solution

Ingest Labs' data collection platform was already deployed at the edge of the firm's web properties, operating as a first-party data collection layer. Because the platform was already in the request path, already scaled to traffic volume, and already managing first-party identity, server-to-server personalization became an activation on existing infrastructure.

The platform's identity solution (MPID) sets a first-party, server-side-only cookie with an approximate two-year lifespan — across all browsers, including those with aggressive ITP policies. MPID is passed as a First-Party ID (FPID) to Adobe's Edge Network, anchoring Target's traffic allocation. Same visitor, same experience, every visit.

The entire personalization flow operates before the page reaches the browser: the visitor request arrives, MPID is resolved, Adobe Target is called server-to-server, and personalization decisions are injected into the response. The visitor's browser never makes a personalization request, never receives raw experiment data, and never has access to the test logic.

A two-datastream architecture preserves Analytics for Target (A4T) reporting without modification — the server-side datastream fetches Target decisions only, while the client-side datastream sends the analytics event with a display notification that stitches the decision back to Adobe Analytics.

Results

Metric Detail
Time to production Days from initiation to live server-side pricing experiments
Infrastructure provisioned None — activated on existing deployment
Identity durability ~2 year cookie lifespan vs. 7-day ITP limit
Security posture Pricing data never exposed to client-side JavaScript
Vendor migration impact Zero — AT.js to Web SDK migration absorbed by platform
Analytics continuity Full A4T reporting preserved via two-datastream architecture
Visitor consistency Deterministic test bucketing via MPID/FPID — same visitor, same price, every visit

Beyond a single experiment

What was built is not an experiment — it is a server-side personalization capability that scales horizontally. Adding a new experiment requires no infrastructure work, no new deployments, and no engineering coordination. The marginal cost of each additional experiment is configuration, not engineering.

The same pipeline that delivers a pricing test can simultaneously deliver homepage content variations, promotional offers, feature flags, or any other JSON-based personalization — all through the same request, the same identity, and the same analytics pipeline.

An enterprise unlocked a security-sensitive personalization use case in days, with no new infrastructure, no vendor lock-in, and full measurement continuity — and gained a platform that scales with their ambitions.

Get started

Ready to run server-side experiments?

Tamper-proof personalization, durable identity, and full analytics — live in days.

Live in <24 hours No code changes SOC 2 compliant