Clayface · S1 scene player · Bloomingdale's demo All figures [sample]
01
Scene 1 · Setting the stage

One week at Bloomingdale's — the world before the story

Clayface is an AI-native market-intelligence platform: it watches every governed KPI × scope against event-adjusted expectations and dispositions everything that moves — so leadership attention only goes where it pays. This demo runs on a Bloomingdale's tenant: real org shape, governed metrics, synthetic [sample] numbers. Across the last few weeks five things moved. The queue sorted all five — and this deck shows the sort, then follows #1 end-to-end.

"As an organizational leader, I want to identify and prioritize anomalous performance signals versus normal variation so I can focus attention on the most critical issues requiring action."
BLOOMINGDALE'S · HERO USE CASE · VERBATIMproved on exactly four sources: 🖱️ summarized clickstream📅 marketing calendar🧾 order data📋 WBR
🏬

The estate

  • 31 full-line doors incl. 59th Street flagship
  • 4 Bloomie's · 25 outlets
  • Dotcom + app — the digital channel this story lives in
  • Regions: Northeast · Southeast · West
📈

The business

  • Banner digital demand ≈ $48.2M/wk
  • Women's RTW ≈ 100K orders/wkthe growth FOB (+7.4% comps)
  • AOV ≈ $200 · contribution ≈ 34%
  • All figures [sample], internally consistent
🗓️

In flight this week

  • Summer Glow sitewide 20–25% off, W27–28 · 6 comparable past promos
  • Beauty $25-off-$250 GWP · Loyallist points event
  • Carousel "Surfer Chic" drop — first-of-kind
  • Site release R-1142, mobile checkout, W26 REMEMBER THIS
🔌

Wired-in sources

  • Summarized clickstream — funnel by device × region
  • Marketing calendar — events + their own history
  • Order data — demand, fulfillment, AUR
  • WBR — certified actuals + prior beliefs. Nothing else assumed.
Where the story starts
W26 · quiet

Release R-1142 ships to mobile checkout. Nobody thinks about it twice.

W27

Summer Glow begins. Traffic climbs — a promo week like six before it.

W28 · reporting wk

Data certified through Sun Jul 05. Somewhere in the noise, a number lands wrong.

MON W29 · 7:00AM

You are here. The WBR is out, the trading call is at 9:30 — and the week must explain itself.

You'll live this week through ACAdrienne Cole · EVP Digital & Omni · leader seat PSPriya Shah · Sr Business Analyst · builder seat Adrienne opens the queue
🎤

For a cold audience — 60 seconds here"Name the tenant honestly: Bloomingdale's shape, synthetic [sample] numbers, the pilot's exact four sources. Then make one promise: by the end of this week, Monday's argument becomes Monday's action."

Scene 2 · The Monday queue

Five things moved. One screen sorts them — before the first meeting.

In the old world, every one of these becomes a fire drill: a dip to investigate, a spike to explain, a drift nobody sees, a pipe that impersonates a crisis. Clayface gives each a verdict with evidence — real anomaly, normal variation, or not a business problem at all. Click any card. All figures [sample].

⚡ Requires attention · ranked by value at stake 2
click
#1

Women's RTW digital — breaks below the promo-adjusted envelope

priority 0.841
−20.3K orders · z −4.1$1.4M/wk at stakepersistence 3 wks

Broke below what six comparable promos support — the promo can't excuse it; concentration named: mobile × Northeast × delivery-promise step.

Old worldA war-room forms around the wrong suspect (the release). Truth takes 2 analysts × 3 days while margin bleeds.
With ClayfaceDriver named, release cleared, action drafted before 9:30. This deck follows this signal end-to-end →
click
#2

Bloomie's + dotcom Premium Contemporary — breaks above the envelope

priority 0.722
+38% · z +3.3+$0.9M/wk above planWOS 2.1 — sizes break ≈W30

Carousel "Surfer Chic" + designer drop is first-of-kind — zero comparables exist, so nothing may suppress it. The envelope flags upside with the same rigor as downside; full-price sell-through 84%, AUR +6% — real demand, not promo froth.

Old worldGood news gets a screenshot in Slack. Sizes break by W30 and the upside quietly caps itself.
With ClayfaceChase-vs-transfer decision teed to the Contemporary DMM with a two-week runway — a growth decision, not a post-mortem.
👁 Watch · analyst validating 1
click

AQUA intimates AQ-1147 — return-rate drift, 4 weeks building

0.495
18.0 → 21.4% · z +1.9$85K/wk · below leader gate

The trap: the sales line looks fine — returns reason-mix shifts to fit/size (41→58%), concentrated in 34B/34C. WBR structurally can't see family-grain returns. Projected to cross the gate ≈ W31; buyer asked to confirm a supplier change.

Old worldDiscovered at season end, after reviews sour — a $1.0M lesson.
With ClayfaceWatched with a named owner and an escalation date — the leader sees one line: "1 emerging · validating."
✓ Expected · suppressed with evidence 1
click

Beauty +32% — the GWP event doing what it always does

z +0.3 adj
inside ×[1.22–1.41] · 6 comparables

The same math that flagged #1 suppressed this — six comparable events say +22…+41% is normal. Suppression is the model working in both directions, and it's auditable: the tray keeps the evidence.

Old world"What happened in Beauty?! Pull it apart by brand" — half a day to prove the obvious.
With ClayfaceZero attention spent; the event's actuals join the comparable set for next time.
⛨ Data health · never reaches the business queue 1
click
!

Mobile-app clickstream lands 8% light

vol z −3.2
routed to feed ownerorders intact · Δ 0.00% to WBR

The paired artifact: traffic "falls" and conversion "jumps" simultaneously — a source symptom, not two business signals. Intact orders are the tell. Affected tiles get a "partial data" chip; the gap is excluded from model training.

Old world"App traffic down 8%!" war-room; four hours later it's a tagging gap — and the next real alert gets doubted.
With ClayfaceQuarantined at the source-health gate before any business disposition was computed. Trust preserved.

The old Monday · 5 spikes = 5 fire drills

  • 8 analyst-days assembling extracts & rebutting each other
  • 1 false accusation — engineering benched to defend a healthy release
  • 1 stockout discovered after the moment passed · 1 fit issue found a season late
  • And the next real alert? Doubted.

The Clayface Monday · attention goes where it pays

  • 2 decisions on the table by 9:30 — one fix, one chase
  • 1 watch with a named owner & escalation date · 1 spike explained with evidence
  • 1 pipe quarantined before it lied to anyone
  • False positives chased: zero.
🎤

The line that lands"Notice what's not here: no hand-tuned thresholds, no alert fatigue. The same envelope logic ran both directions in the same week — it suppressed Beauty's +32% and flagged RTW's dip. That's 'anomaly versus normal variation,' proven."

REQUIRES ATTENTION SIGNAL 1 OF 5 · QUEUE RANK #1 Rank #1 · priority 0.841 Bloomingdale's tenant · W28 FY2026

The Monday the promo hid a $1.4M/wk problem — and Clayface found it before coffee.

Mid–Summer Glow promo, Women's RTW digital orders quietly land 7.5K below the lower bound of what six comparable promos support. The loss hides inside promo noise, concentrates in mobile checkout in the Northeast, and traces to a carrier disruption stretching delivery promises from 2-day to 5–7 days. A site release looks guilty — and is exonerated by the evidence.

#1 Women's RTW · digital demand breaks below the event-adjusted envelope queue: requires-attention · maker/checker ✓
W28 actual
0
K orders · vs plan ≈100K
Event-adjusted envelope
[98.7 · 111.5 · 125.1]
K · from 6 comparable promos
Surprise
z = −4.1
>99.9th pctile · persistence 3 wks
Value at stake
$0M
contribution margin / week
Priority
0
blm-demo-priority-v1
The cast ACAdrienne Cole · EVP Digital & Omni · leader (default seat) PSPriya Shah · Sr Business Analyst (EDA) · builder seat DCDigital Commerce lead · action owner · data object SDS. Delgado · VP Fulfillment · action owner · data object
🎤

Open with"This is the required use case, verbatim: separate anomalous signals from normal variation — during the noisiest possible week to do it: a sitewide promotion." Then tap the Priority tile to show the math is governed, not vibes.

Scene 4 · The business moment

Monday 7:00am, Week 29. Same data. Two very different mornings.

The WBR shows Women's RTW digital soft vs. plan — but the Summer Glow promo gives everyone an excuse to argue. Click each door to see where the morning ends up; the quotes arrive as you advance.

click → outcome
WITHOUT CLAYFACE · MON 7:00 AM

The argument forms before the facts do

Soft number, promo week, five dashboards, zero shared baseline. Every function reads its own tool and defends its own turf.

"It's just promo mix — AOV always dips." "Traffic looks fine, sessions are +9%." "…didn't engineering ship a release two weeks ago?"
Outcome → a war-room forms around the wrong suspect. The truth takes 2 analysts × 3 days.
click → outcome
WITH CLAYFACE · MON 7:00 AM

The queue's #1 item already settled it

The week landed below what comparable promos support. Concentration named: mobile × Northeast × the delivery-promise step. Strongest supported driver: carrier disruption. R-1142? Evaluated and cleared.

One card. Evidence attached. Nothing to argue about.
Outcome → the 9:30 trading call starts at the fix, not the blame.
🎤

Framing"Keep both doors on screen. The room has lived the left one — let them recognize it before you show the race."

Scene 5 · Side by side

The war-room week vs. the 30-minute read — press play

One clock, two organizations. Watch the margin meter on the right while the left side argues. Drag the scrubber to any moment.

MON · 7:00
MON 7:00MON 9:30MON 11:00TUEWED–THUFRI 16:00
Margin exposure
while arguing
$0
🔥 WITHOUT CLAYFACE

MON 7:00WBR lands: RTW digital soft vs. plan SIGNAL BURIED

One row in a 40-tab workbook. The promo makes the miss deniable.

MON 9:30Trading call becomes a debate club NO SHARED BASELINE

Marketing cites sessions +9%, merchandising blames promo mix, finance sees an 18% gap. Three tools, three truths.

MON 11:00War-room forms — around the wrong suspect FALSE ACCUSATION

"Didn't engineering ship a release?" R-1142 becomes prime suspect; tech benched to defend itself.

TUETwo analysts start hand-stitching four systems 2 ANALYSTS PULLED

Order data, clickstream exports, promo calendar, WBR files — reconciling definitions before analyzing.

WED–THUStill reconciling DAY 3–4

Every function re-cuts the data to defend its own story. The miss keeps compounding.

FRI 16:00The truth finally surfaces 4 DAYS LATE

It was never the release — a carrier disruption stretched NE promises to 5–7 days. Mitigation starts next week.

WITH CLAYFACE

MON 7:00Queue #1: card with everything DETECTED

Breach vs the event-adjusted envelope · $1.4M/wk at stake · confidence High · maker/checker ✓.

MON 7:05The "why" is attached EXPLAINED

Mobile × NE −11.2K · delivery-option step · promise slip 2d → 5–7d · R-1142 evaluated & cleared.

MON 9:30Trading call starts at the fix NO WAR ROOM

Priya validates method in the evidence drawer instead of rebuilding it. Tech never gets accused.

MON EODAction object assigned, due 48h ASSIGNED

fulfillment-mitigation → Digital Commerce lead + Delgado · ship-from-store, pickup at checkout, Loyallist expedite, honest promises.

✓ Done since Monday. Engineers on roadmap. Analysts analyzing.
Mitigation already recovering the funnel.

3 days~30 min
Signal → validated driver
2 analysts0
Headcount spent assembling data
War roomCleared
Tech exonerated with evidence
≈$0.9M/wk
Recovered by W30 · [sample]
🎤

Land this line"Detect → explain → assign → approve → measure. And notice the platform spent its first breath clearing your tech team, not accusing it."

Scene 6 · Detection

The envelope that promos can't fool

Not a static plan — an event-adjusted envelope trained on the six comparable promos in the calendar's own history. Hover the points; flip the toggle to see why the old way sleeps through this week.

Women's RTW · digital orders/wk · W17–W28 [sample]

envelope actual breach
old static-threshold world
The old way never fires: "alert at plan −10%" sits at 90K — W28's 91.2K glides right over it. Meanwhile sessions are +9%, so "traffic looks fine" wins the meeting. The envelope fires because the promo should have delivered 98.7–125.1K.
156-wk training 6 comparables → uplift [+5%, +18%] σ̂ ≈ 4.96K deterministic · reproducible digital_demand_sales · def v3.2
🎤

The differentiator"A promotion is precisely when static thresholds go blind — the miss is measured against the event's own history, so the event cannot excuse the miss."

Scene 7 · The four-source proof

No single source sees this. All four together can't miss it.

Click each source for its contribution — then drill the decomposition to find where the orders went.

+ click
📅

Marketing calendar

Summer Glow W27–28 · 6 comparables → envelope lifted to [98.7, 125.1]K.

Kills the "it's just the promo" excuse
+ click
🧾

Order data

Demand −18% vs adjusted median · AUR flat · BOPIS +3.1pp in NE doors while ship-to-home falls.

Locates the loss · rules out markdown/mix
+ click
🖱️

Summarized clickstream

Sessions +9% — promo working! Funnel healthy until mobile completion −7.8pp at the delivery-option step.

Names the exact broken step — from summaries, not raw logs
+ click
📋

WBR data

W27 disposition "monitor — promo noise expected" captured · W28 certified actuals reconcile at Δ 0.00%.

A business problem, not a data problem (contrast: S4)

Where did −20.3K orders go?

Decomposition vs event-adjusted median · % = vs each segment's own adjusted expectation · sums exactly · [sample] · click a bar to drill
Mobile × Northeast
−11.2K−33%
Mobile × other regions
−4.6K−11%
Desktop
−2.1K−9%
Store-fulfilled digital
−1.4K−11%
Mix
−1.0Kresid.
Total vs adjusted median−20.3K · −18.2%
The anomaly in one glance: Mobile × NE is off −33% vs its own promo-adjusted expectation — 3× worse than every other segment (−9…−11%). Concentration this sharp is a localized break, not soft demand.
Class cut is diffuse (dresses −19% · denim −17% · tops −18%) — diffuseness is itself evidence for a funnel driver, not a merchandise driver. That's why the signal correctly fired at FOB grain.
🔍Drill Mobile × Northeast on the left
to open the funnel view

The funnel is healthy — until one step

Mobile × Northeast vs expected · W28 · [sample]
Sessions
+9%
PDP views
+7%
Add-to-bag
+5%
Checkout starts
−1.2%
Completion
−7.8pp
2-day5–7 days

Delivery-option step: NE promise slipped mid-W27. Mobile completion 87.4% → 79.6%; desktop only −1.2pp.

Cross-channel signature: shoppers who could switch to pickup did (BOPIS +3.1pp); those who couldn't, abandoned. No single-source tool sees both halves.
−20.3K orders× $200 AOV× ~34% contribution $1.4M / week The value model is shown, never implicit — evaluators can recompute the headline themselves. [sample]
Scene 8 · Competing explanations

The suspect board — listed, not hidden

The demo's political high point: the platform evaluates every plausible explanation including the one the room would jump to — and shows its work. Click each suspect for the verdict stamp.

🚚

Carrier network disruption — NE lanes

External operational event · mid-W27 onward
REVEAL
STRONGEST SUPPORTED DRIVER
  • Promise display slipped 2-day → 5–7 days in NE ship-to-home zips, exactly at breach onset (mid-W27)
  • Region-specific and device-skewed — promo traffic skews toward promise-sensitive shoppers
  • Corroborated cross-channel: BOPIS +3.1pp in NE doors while ship-to-home falls
  • External cause — nobody in the room built this problem
🚀

Site release R-1142 — mobile checkout

Shipped W26 · the room's most tempting accusation
REVEAL
EVALUATED & CLEARED ✓
  • Release is platform-wide; the impact is region-specific — a defect wouldn't respect geography
  • Timing misaligns: release W26, breach onset mid-W27 — at the promise slip, not the deploy
  • Desktop shows only mild promise-sensitivity (−1.2pp), not a defect signature
  • The platform clears the tech team with evidence. No war room. No benched engineers.
🎟️

Loyallist promo-stacking friction

Points event overlapping Summer Glow exclusions
REVEAL
CANNOT RULE OUT · LOW WEIGHT
  • Plausible during overlapping events — but would not be region-specific
  • Kept on the list honestly; the platform never pretends to certainty it doesn't have
🏬

Competitor counter-promo in NE

Outside the four pilot sources
REVEAL
OUTSIDE ESTATE · ROADMAP
  • Needs open-market signal — a Layer 5 source, honestly labeled as roadmap
  • Named validation instead of fabricated confidence — the trust posture the platform is built on
Language discipline (enforced in-product): "The carrier caused the miss" "The strongest supported driver is… validate via carrier SLA reports, NE promise-display values, ship-from-store capacity at the 12 NE doors."
🎤

For Miller (Product & Tech Strategy)"A tool that ends 'blame engineering first' war-rooms is a tool Product & Tech Strategy sponsors. Your release got accused at 11am and acquitted by the data before lunch."

Scene 9 · Act & measure

From verdict to action object — and a ledger that keeps score

The signal becomes an owned, approved, deadline-bearing action. Approve it as Adrienne — then watch the ledger close the loop. Owner ≠ approver, by design: approval can't be bypassed.

action object · fulfillment-mitigation · label: demo-workflow

Recover the NE mobile checkout gap

Joint owners
Digital Commerce lead + S. Delgado (Fulfillment)
Approver
Adrienne Cole (EVP Digital & Omni)
Due
48 hours
Expected outcome
Recover ≥60% of completion gap by W30
01Activate priority ship-from-store from the 12 NE doors — restore fast promises with in-region inventory
02Surface pickup / same-day options at the delivery step — give promise-sensitive shoppers an out
03Expedite-upgrade Loyallist orders — protect the highest-value cohort first
04Adjust promise messaging — honest dates beat optimistic ones that drive abandonment
Honest label: in the demo this is a demo-workflow; target state hands off to order-routing / fulfillment. Rationale pre-drafted; every state change audited.

The ledger · closing the loop

Every claim gets measured — and caveated · [sample]
W28

Signal fires. Breach z = −4.1 · $1.4M/wk at stake · action approved same day.

W29

Mitigation live. Ship-from-store routing on; pickup surfaced at checkout; NE promises truthful.

W30

Completion recovered to −0.9pp vs expected · $0.0M/wk recovered.
Attribution caveat recorded: "carrier lanes partially restored in parallel; recovery not solely attributable."

Strategic footnote: NE ship-from-store share holds +6pp after the disruption clears — the mitigation became a structural omnichannel win. That's the story Fielder retells.
🎤

Trust beat"Read the caveat out loud. A platform that hedges its own success honestly is a platform you can believe when it claims a $1.4M problem."

Scene 10 · The weeks after

Three weeks later: all five, closed with receipts

Signals don't end at insight. Each got an owner, a decision, and a measured outcome in the ledger — with caveats recorded, because trust compounds. All figures [sample].

#1W28 → W30

RTW digital fix live in 48h. Completion recovered to −0.9pp · ≈$0.9M/wk back · ship-from-store share holds +6pp after the disruption clears — a structural omnichannel win.Caveat recorded: carrier lanes partially restored in parallel; recovery not solely attributable.

#2W28 → W31

Chase + transfer approved Monday. +$2.1M incremental demand captured vs do-nothing · sizes held through W30 · rotation window extended.Caveat recorded: rotation extended and chase landed — effects not separable.

W29

Buyer confirms the supplier change on AQ-1147 POs → "runs small" PDP nudge + size-curve correction on next receipt. Caught ≈12 weeks before season-end discovery — ≈$1.0M margin difference.

Beauty stayed suppressed. Zero analyst-days spent proving the obvious; the event's actuals joined the comparable set — next quarter's suppression just got smarter.

!W28 THU

Feed backfilled, checks green, chip cleared. The gap stays excluded from model training, so the model never learns that low traffic is "normal" — and the next real alert arrives untarnished.

≈$0.9M/wkmargin protected
+$2.1Mupside captured
≈$1.0Msaved by early detection
0analyst-days on false positives
2leadership decisions — the whole point
🎤

Trust beat"Every number on this screen sits next to its caveat. A platform that hedges its own wins honestly is a platform you can believe on a Monday morning."

Scene 11 · The new day-to-day

A new era: agentic intelligence, human decisions

This isn't a dashboard upgrade. It's a different operating rhythm for Bloomingdale's: agents do the watching, assembling, and explaining — continuously — so the leader and the teams spend their week deciding and acting, not proving and arguing.

✕ The old rhythm✓ With Clayface
MON 7:00

Dashboard archaeology: 40 tabs, five tools, three versions of the truth

One ranked queue, dispositioned with evidence — reading, not assembling

MON 9:30

The meeting argues about whether it's real

The meeting starts at the fix and the chase

TUE–THU

Analysts hand-stitch extracts; engineering defends a healthy release

Analysts validate and extend models; engineering ships roadmap

FRI

The truth arrives four days late; the good news already capped itself

The ledger reports outcomes with caveats — and the models learn the week

The agentic loop · running every week, on every governed KPI × scope
SENSESevery KPI × scope, event-adjusted EXPLAINSfour-source evidence, suspects cleared PRIORITIZESby value at stake, deterministic ROUTESnamed owners, approval gates MEASURESledger, caveats recorded LEARNSevery event joins the history
Agents do the assembling. People make the calls.
continuous · deterministic · governed · audited · [sample] world
L

For the leader

Adrienne's seat

Attention becomes the optimized resource. Monday opens on decision #1 and #2 — not on "is it real?". Five moves, two meetings, zero chased ghosts.

3 days of arguing → 30 minutes of deciding
T

For merchant, planning & fulfillment

the owners

Signals arrive at their grain — style × size, door, funnel step — with owners, deadlines, and runway: chase while sizes last, fix while the promo still runs.

Decisions on the timeline where they still matter
D

For tech, data & analytics

the builders

A platform that defends its own rigor — clears releases with evidence, quarantines bad pipes before they lie, and treats feed owners as first-class users.

No more war-rooms that end in "it was tagging"

"Five things moved. Two were explained away with evidence, one was quarantined, one went on watch with a named owner — and leadership attention was spent only on the two that deserved it. That discipline, every single week, is the new era."

Clayface — agentic market intelligence for Bloomingdale's. Questions welcome.

Why this rank? · priority 0.841

Weights blm-demo-priority-v1 · severity 0.30 · impact 0.35 · confidence 0.20 · actionability 0.15 · [sample]
Severity
0.845× .30
↳ surprise 0.95 · persistence 0.60
.7/.3
Impact
0.90× .35
Confidence
0.80× .20
Actionability
0.75× .15
Weighted priority0.841
Deterministic and reproducible — the same inputs always produce the same rank. That's the trust posture the platform is built on.