FullStory for El Palacio de Hierro — Unified Experience Analytics

The Diagnosis

You're paying for behavioral analytics twice

Every event your team wants to track today is instrumented in two places — once in GTM (where it's wired into the GA4-compatible Enhanced Ecommerce schema) and a second time in Quantum Metric (where the implementation team writes CSS selectors). Two scopes, two ticket queues, two sets of breakage when the front end ships.

17

Events configured in QM
vs ~92 typical for retailer of this scale

2

CSS selectors used by QM
(.l-plp, .l-plp .b-breadcrumbs-link)

0

PII scrubbing rules configured
compliance gap under LFPDPPP

0

Cart, checkout, purchase, revenue
not tracked in QM today

Today · Quantum Metric

The instrumentation tax — paid twice, coverage is partial

QM is technically deployed but functionally a stub. The team would still need to write CSS selectors for every event they ever want to capture, and as the front-end ships changes those selectors break. Microsoft Clarity is also active, duplicating QM's session-replay capability.

No data-layer integration — DOM/XHR scraping only
Cart, checkout, purchase events not tracked
Custom events (interaccion_search_ai, filter_interaction) ignored
Zero PII scrubbing rules — compliance exposure
Microsoft Clarity adds a third replay tool with the same gaps

With FullStory

Instrument once. Distribute everywhere.

FullStory ingests your existing GTM data layer via DLO with zero developer work. FullCapture records every session by default. Activation Streams pushes events back to GA4, Google Ads, and other destinations — your team configures one event in FullStory; it lands in every reporting destination automatically.

DLO consumes your existing Enhanced Ecommerce schema natively
Cart, checkout, purchase, revenue captured day one
Custom events flow in automatically — no rule writing
Default PII catalog (CVV, RFC, CURP, card numbers, email)
Replaces both QM and Clarity

Your real data layer

You're not data-poor. You're data-fragmented.

We captured these payloads live from elpalaciodehierro.com on 2026-04-30 — a real PDP, a real PLP, with the QM script running on the page the entire time. Every field below is data your team built into GTM. Quantum Metric is reading none of it.

Two FullStory sessions were recorded while QM was active on the page — one on the Tissot PRX PDP, one on the Burberry brand PLP. The contrast between what FS captured and what QM's 17 events captured is its own demo. Live session links in the appendix.

PDP · productDetail event

// Tissot PRX Casual Plateado, $7,245 MXN
{
  "event": "productDetail",
  "ecommerce": { "detail": { "products": [{
    "id": "42316607",
    "name": "Reloj para Hombre PRX casual plateado",
    "price": 7245,
    "brand": "Tissot",
    "brandID": "E704",
    "category": "RELOJERIA",
    "departmentName": "RELOJERIA",
    "departmentID": "50401",
    "subClase": "504010210",
    "dimension18": "RELOJERIA JOYERIA Y ESCRITURA",
    "dimension19": "RELOJ CABALLERO"
  }] } }
}

PLP · productImpression event (Burberry, 787 products)

// Position-ranked, AI-personalized via Constructor.io
{
  "event": "productImpression",
  "constructor_activity": true,
  "ecommerce": {
    "currencyCode": "MXN",
    "impressions": [{
      "id": "44926644",
      "name": "Bolsa Freya",
      "price": 38990,
      "brand": "Burberry",
      "category": "BOUTIQUES LUJO",
      "dimension18": "LUJO MODA",
      "list": "Resultado Busqueda",
      "position": 10
    }]
  }
}

The luxury angle that the current stack misses: every product impression carries position (rank in the AI-ranked list) and list (source attribution). This is the data needed to answer "does Constructor.io's #3 ranking convert better than #10 on $38K Burberry handbags?" — a question worth real revenue at luxury price points. QM has no productImpression instrumentation; FullStory ingests it natively via DLO.

Custom events your team built that no system is currently consuming

Custom GTM Event

interaccion_search_ai

AI search engagement, fired by Constructor.io's discovery layer. Critical for measuring how AI search converts on luxury inventory.

Custom GTM Event

filter_interaction

Filter sidebar usage on PLPs. Tells you what shoppers narrow down before they buy. Today: instrumented but unused downstream.

Standard EE Event

promotionImpression

Promo creative views with id, name, creative, position. Measures which Noches Palacio creatives drive carts.

The Architecture

Capture once. Ingest natively. Distribute everywhere.

The shift is not "add another tool." It's the opposite — collapse three replay tools into one, eliminate duplicate GTM-vs-QM instrumentation, and use FullStory as the single source of truth for the user experience that flows out to every system that needs it.

Today · Fragmented

Three behavioral tools, partial coverage from each

GTM dataLayer rich and well-modeled — but only consumed by GA4 + ad pixels
Quantum Metric: 17 events, 2 selectors, no DLO, no PII rules
Microsoft Clarity: duplicates QM session replay
FullStory: already partially loaded, not orchestrated
Custom events (interaccion_search_ai, filter_interaction) have no destination reading them
Every new event = two implementation tickets (GTM + QM)

Proposed · Unified on FullStory

One source, every destination

FullCapture records every session by default
DLO ingests existing GTM dataLayer natively
Activation Streams pushes events to GA4, Google Ads, Floodlight, Meta, TikTok
Warehouse Sync delivers structured behavioral data to BigQuery / Snowflake
Salesforce CDP gets behavioral signals (rage clicks, frustration, abandon) as attributes
One instrumentation source. One implementation ticket.

BEFORE AFTER ──────────────────── ────────────────────────────── Site (SFCC) Site (SFCC) │ │ ├─► GTM ──┬─► GA4 ├─► GTM dataLayer ──┐ │ ├─► 8× Google Ads │ │ │ └─► Floodlight ├─► Ad pixels (kept) │ │ │ │ ├─► Quantum Metric (17 evt) └─► FullStory ◄───────────┘ via DLO ├─► Microsoft Clarity │ ├─► FullStory (partial) ├─► FullCapture (all sessions) ├─► CQuotient + MC Pers │ ├─► Constructor.io ├─► Activation Streams ─┬─► GA4 ├─► Bazaarvoice │ ├─► Google Ads ├─► Datadog RUM │ ├─► Meta / TikTok └─► dw.ac (SFCC native) │ └─► Floodlight │ └─► Warehouse Sync ─────┬─► BigQuery ├─► Snowflake └─► Salesforce CDP Decommission: Quantum Metric, Microsoft Clarity Keep & integrate: GTM, dw.ac, CQuotient, MC Personalization, Constructor.io, Bazaarvoice, Datadog

What FullStory delivers

Day-one capabilities, mapped to your stack

These are the four products that solve the specific problems we identified in your current stack. Each one maps directly to a gap that QM, Clarity, or the duplicate-instrumentation pattern is creating today.

FullCapture + DLO

Capture everything. Read your data layer.

✓

Every session by default

Clicks, scrolls, network requests, console errors, form interactions, frustration signals.

✓

Native GTM dataLayer ingestion

Enhanced Ecommerce v3 + GA4 schema recognized out of the box. Custom events via JSON rules.

✓

Frustration signals built-in

Rage clicks, dead clicks, error clicks, form rage. No rules to write.

Activation Streams

Instrument once. Distribute everywhere.

→

Real-time event streams to GA4

Configure the event in FullStory; it lands in GA4. No more double-instrumenting GTM and QM.

→

Direct to ad platforms

Google Ads, Floodlight, Meta, TikTok, Pinterest — all from one source.

→

Frustration signals as audiences

"Users who rage-clicked the cart button on Burberry" becomes a Google Ads remarketing audience.

Warehouse Sync

Unify CDP + data science around the user.

⊕

BigQuery / Snowflake delivery

Structured event tables, partitioned by date / org / event. Joinable to SFCC orders, Constructor logs, ad conversions.

⊕

Salesforce CDP enablement

Behavioral attributes (cart abandon, frustration, AI-search engagement) flow into your CDP for unified profiles.

⊕

Data-science feature store

LTV, churn, propensity-to-convert models can finally use behavioral features, not just transactional ones.

Privacy & compliance

Default-on PII protection.

⊘

Default PII catalog

CVV, RFC, CURP, card numbers, email, phone — masked or excluded out of the box. QM today: zero PII rules configured.

⊘

Field-level masking

fs-mask / fs-exclude attributes preserve replay while hiding the input.

⊘

LFPDPPP-aligned

Regional controls aligned with Mexican data-protection requirements. Console redaction, IP exclusion, URL exclusion.

The killer feature for your situation

Activation Streams ends the duplicate-instrumentation problem

Today, every event your team wants to track gets instrumented in GTM, then re-instrumented in QM. Activation Streams replaces that pattern: configure the event once in FullStory; it lands in GA4, Google Ads, Floodlight, Meta, and any other destination automatically.

Event Palacio cares about	Captured by	Streamed automatically to
add_to_cart	DLO ingesting GTM addToCart	GA4, Google Ads, Meta Pixel, Floodlight
purchase (when implemented)	DLO ingesting GA4 purchase	GA4, Google Ads, Floodlight, BigQuery
product_impression with position	DLO ingesting productImpression	BigQuery (for AI-rank ROI analysis)
interaccion_search_ai	DLO ingesting custom event	GA4, BigQuery
filter_interaction	DLO ingesting custom event	GA4, BigQuery
rage_click_on_cart_button	FullCapture auto-detected	GA4 (as anomaly), Slack, BigQuery
checkout_frustration_signal	FullCapture auto-detected	GA4, Salesforce CDP, retargeting audiences

What this enables that isn't possible today: streaming behavioral signals (rage clicks, dead clicks, frustration scores) into GA4 and ad platforms as audiences. Today these signals don't exist anywhere in Palacio's stack — QM nominally captures them but doesn't expose the data, and GA4 has no native concept of frustration. With FullStory + Activation Streams, "users who rage-clicked the cart button on Burberry handbags this week" becomes a Google Ads remarketing audience.

Concrete deliverable

DLO rules ready to deploy against your data layer

These rules were generated from your actual window.dataLayer schema, captured live on 2026-04-30. They map the events your GTM is already firing into normalized FullStory events that downstream Activation Streams and warehouse exports can use cleanly. No developer change to your site is required to deploy them.

FullStory DLO rules · drop into the DLO config UI

[
  // ─── Page-level baseline ───
  {
    "id": "page_view",
    "source": "dataLayer",
    "match": { "pageType": { "$exists": true } },
    "destination": "FS.event",
    "fsEvent": "Page Viewed",
    "properties": ["pageType", "pageTitle", "userId1", "userId2", "userStatus"]
  },

  // ─── Product Detail Page view ───
  {
    "id": "product_view",
    "source": "dataLayer",
    "match": { "event": "productDetail" },
    "destination": "FS.event",
    "fsEvent": "Product Viewed",
    "renameMap": {
      "ecommerce.detail.products[0].id": "product_id",
      "ecommerce.detail.products[0].name": "product_name",
      "ecommerce.detail.products[0].price": "price",
      "ecommerce.detail.products[0].brand": "brand",
      "ecommerce.detail.products[0].category": "category",
      "ecommerce.detail.products[0].departmentName": "department",
      "ecommerce.detail.products[0].dimension18": "segment_lujo",
      "ecommerce.detail.products[0].subClase": "subclass"
    }
  },

  // ─── Add to Cart ───
  {
    "id": "add_to_cart",
    "match": { "event": "addToCart" },
    "fsEvent": "Add To Cart",
    "renameMap": {
      "ecommerce.add.products[0].id": "product_id",
      "ecommerce.add.products[0].price": "price",
      "ecommerce.add.products[0].quantity": "quantity",
      "ecommerce.add.products[0].brand": "brand"
    }
  },

  // ─── PLP impression (fan-out per item, preserves position + list) ───
  {
    "id": "product_impression",
    "match": { "event": "productImpression" },
    "fsEvent": "Product Impression",
    "fanOutOver": "ecommerce.impressions",
    "renameMap": {
      "id": "product_id",
      "price": "price",
      "brand": "brand",
      "list": "list_source",
      "position": "list_position"
    }
  },

  // ─── AI search interaction ───
  { "id": "ai_search", "match": { "event": "interaccion_search_ai" },
    "fsEvent": "AI Search Interaction" },

  // ─── Filter interaction ───
  { "id": "filter", "match": { "event": "filter_interaction" },
    "fsEvent": "Filter Used" },

  // ─── Purchase (when checkout instrumentation is added) ───
  {
    "id": "purchase",
    "match": { "event": "purchase" },
    "fsEvent": "Order Completed",
    "renameMap": {
      "ecommerce.purchase.actionField.id": "order_id",
      "ecommerce.purchase.actionField.revenue": "revenue",
      "ecommerce.purchase.actionField.tax": "tax",
      "ecommerce.purchase.actionField.shipping": "shipping",
      "ecommerce.currencyCode": "currency"
    }
  }
]

Stack consolidation map

Keep, replace, decommission

Most of your stack is doing useful work. The redundancy is concentrated in behavioral analytics and session replay, where three tools overlap. The map below classifies each system by recommended action.

System	Today's role	Recommendation	Why
Behavioral analytics & replay
Quantum Metric	Replay + 17 events	Decommission	Stub deployment, no DLO, doubles instrumentation cost
Microsoft Clarity	Replay only	Decommission	Duplicates QM, less mature than FS
FullStory	Partial deployment	Promote to primary	Replaces both above; ingests existing data layer via DLO
Tag management & analytics
Google Tag Manager	Tag orchestration	Keep	Source of truth for the dataLayer FS DLO consumes
Google Analytics 4	Reporting	Fed via Activation Streams	No more double-instrumentation; FS streams events in
Google Ads (8 containers)	Conversion + retargeting	Fed via Activation Streams	FS frustration signals add new audiences
Floodlight / DoubleClick	Conversion measurement	Fed via Activation Streams	No change to ad-platform reporting model
Meta / TikTok / Pinterest / Bing UET / RTB House	Pixel-based ads	Fed via Activation Streams	One source streams to all
Personalization, search, commerce
SFCC native (dw.ac)	First-party SFCC analytics	Keep	Free, native, not duplicative
CQuotient / Einstein	Personalization & recs	Keep + integrate	FS audiences feed Einstein decisions
MC Personalization (igoDigital)	Cross-channel personalization	Keep + integrate	FS frustration signals become CDP attributes
Constructor.io	AI search & discovery	Keep + measure	FS measures conversion impact of Constructor's rankings
Reviews & performance
Bazaarvoice	Reviews UGC	Keep	Different job, not duplicative
Datadog RUM	Performance monitoring	Keep	Different job, not duplicative

Spend redirection. The Quantum Metric and Microsoft Clarity line items become available for FullStory + Activation Streams + Warehouse Sync. The net P&L impact depends on Palacio's contracts; we can build the side-by-side once we know the QM and Clarity terms.

30 / 60 / 90 day rollout

Validation in 14 days. Decommission decision in 90.

The plan below is paced to clear the validation hurdle quickly, prove parity with QM in production shadow mode, and reach a decommission decision before peak Q3/Q4 trading season.

1

Days 0–14 · Validation

Stand up & confirm coverage

FullStory eval org against staging
Snippet via existing GTM tag
Apply DLO rule catalog
Validate against 5-session sample

2

Days 15–30 · Production shadow

Run alongside QM

Shadow mode in production
First Activation Stream live: FS → GA4
Frustration alerts to Slack
Exec-level review meeting

3

Days 31–60 · Full coverage

All streams live

Full DLO catalog in production
Streams to GA4, Ads, Floodlight
Warehouse Sync configuration begins
QBR with quantified gap-closure metric

4

Days 61–90 · Consolidate

Warehouse + decommission

Behavioral data flowing to BigQuery / Snowflake
Salesforce CDP integration prototyped
Decommission decision: QM + Clarity
Contractual transition plan

Recommendation

Validate against your own site this week

The clearest evaluation path is a 60-minute working session against elpalaciodehierro.com with your analytics team in the room. We bring the DLO rule catalog, you bring a sandbox URL, and we stand up the FS eval org live. The validation itself is the demo.

Day-One Wins

Frustration signals on the cart-add flow, AI-rank conversion analysis, default PII protection

Three measurable improvements over the current QM build, deliverable inside the validation window. None of them require front-end engineering work from Palacio's team.

The Strategic Win

Activation Streams ends the GTM-vs-QM duplicate-instrumentation tax

One implementation source, every reporting destination. The analytics team stops shipping the same event twice. Going forward, "instrument once" is the operating model — not a slogan.

The 90-day Outcome

FullStory becomes the source of truth for the user experience, distributed everywhere

QM and Clarity decommissioned. GA4 and ad platforms fed automatically. BigQuery / Snowflake unified for CDP and data science. Salesforce CDP receiving real-time behavioral signals. One layer. One truth.

Appendix · Audit evidence

Every claim in this brief is sourced from the live site

The findings here come from direct observation of elpalaciodehierro.com on 2026-04-30 — including a real PDP cart-add flow and the Burberry brand PLP. Both sessions were captured by FullStory itself, in real time, while Quantum Metric was running on the same page. The contrast is its own demo.

Artifact	What it proves
QM SDK snapshot · v1.35.37	17 events, 2 selectors, 0 PII rules, no DLO
Live data layer captures · PDP + PLP	Enhanced Ecommerce v3 + GA4 mirror, custom events, AI-search flag
Vendor stack inventory · 22 tools detected	Three replay tools (QM + Clarity + FS partial), eight Google Ads containers
FullStory session · PDP cart-add flow	End-to-end PDP load → cookie consent → cart-add captured by FS while QM was running
FullStory session · Burberry PLP	787-product PLP impression event captured with full position + list attribution

One source of experience truth.
Distributed everywhere it's needed.

You already built the data layer. Let's make it pay off.

The Diagnosis

Your Data Layer

Unified Architecture

30 / 60 / 90 Plan

The instrumentation tax — paid twice, coverage is partial

Instrument once. Distribute everywhere.

Custom events your team built that no system is currently consuming

interaccion_search_ai

filter_interaction

promotionImpression

Three behavioral tools, partial coverage from each

One source, every destination

Capture everything. Read your data layer.

Every session by default

Native GTM dataLayer ingestion

Frustration signals built-in

Instrument once. Distribute everywhere.

Real-time event streams to GA4

Direct to ad platforms

Frustration signals as audiences

Unify CDP + data science around the user.

BigQuery / Snowflake delivery

Salesforce CDP enablement

Data-science feature store

Default-on PII protection.

Default PII catalog

Field-level masking

LFPDPPP-aligned

Stand up & confirm coverage

Run alongside QM

All streams live

Warehouse + decommission

Validate against your own site this week

Frustration signals on the cart-add flow, AI-rank conversion analysis, default PII protection

Activation Streams ends the GTM-vs-QM duplicate-instrumentation tax

FullStory becomes the source of truth for the user experience, distributed everywhere

One source of experience truth.Distributed everywhere it's needed.

You already built the data layer. Let's make it pay off.

The Diagnosis

Your Data Layer

Unified Architecture

30 / 60 / 90 Plan

The instrumentation tax — paid twice, coverage is partial

Instrument once. Distribute everywhere.

Custom events your team built that no system is currently consuming

interaccion_search_ai

filter_interaction

promotionImpression

Three behavioral tools, partial coverage from each

One source, every destination

Capture everything. Read your data layer.

Every session by default

Native GTM dataLayer ingestion

Frustration signals built-in

Instrument once. Distribute everywhere.

Real-time event streams to GA4

Direct to ad platforms

Frustration signals as audiences

Unify CDP + data science around the user.

BigQuery / Snowflake delivery

Salesforce CDP enablement

Data-science feature store

Default-on PII protection.

Default PII catalog

Field-level masking

LFPDPPP-aligned

Stand up & confirm coverage

Run alongside QM

All streams live

Warehouse + decommission

Validate against your own site this week

Frustration signals on the cart-add flow, AI-rank conversion analysis, default PII protection

Activation Streams ends the GTM-vs-QM duplicate-instrumentation tax

FullStory becomes the source of truth for the user experience, distributed everywhere

One source of experience truth.
Distributed everywhere it's needed.