Value Packs / Retail / Retention & Loyalty Early-Warning / Brief 02

Retention & Loyalty Early-Warning · Brief 02

Cyber Week · a region turned detractor before anything else moved

Southeast · week of 2025-12-01 · pickup reliability → guest sentiment → the ads that depended on it

DataGenie Brief · autonomous
Southeast · bottom-quartile remodel stores · merchandising + marketplace + retail media · illustrative

The one line you configure

Guest sentiment moves before guests do. Which regions are running below their own NPS baseline this week, and is anything operational underneath it?

In a nutshell

One region, and the cause is fourteen stores. Guest NPS across the Southeast sits 18 points below its own baseline30% detractors against 40% promoters, which is an NPS of 10 where 28 is normal. Underneath it, the bottom-quartile remodel stores in the region ran 62% curbside on-time through Cyber Week against a ~97% chain baseline, with stockouts at 1.8×. Guests who clicked through to collect could not collect, and the conversion-objective ads that sent them are converting 26% below plan as a result.

At a glance

Guest NPS · Southeast10vs 28 baseline18 points below
Detractor share30%of respondentsthe leading edge
Promoter share40%capped
Curbside on-time rate62%vs ~97% baselinethe cause
Stockout instances1.8×vs baseline
Conversion-ad attribution−26%Beauty & Household

What moved

Guest NPS · Southeastagainst the region's own baseline of 2810
Curbside on-time · 14 bottom-quartile stores35 points below the rest of the chain62%
Stockout instances · same cohortcompounding the failed collections1.8×
Conversion-objective ads · Beauty & Householdthe click-to-pickup loop is broken−26%

What · the sentiment mix behind an NPS of 10

10Guest NPS · vs 28
Promoterscapped at 40% this week40%
Passivesthe remainder30%
Detractorslifted from the cross-shop base30%

NPS is promoters minus detractors. At 40 and 30 the region scores 10 against its own 28 — and the shape matters more than the score: promoters did not fall away, detractors were created. That is a specific bad experience, not a slow drift in preference.

Why · pickup reliability in fourteen stores, week by week

Curbside On-Time Rate · weekly−35 pts
50%62.5%75%87.5%100%62% — Cyber WeekNov 10Nov 17Nov 24Dec 1Dec 8
Cyber Week62%
Chain baseline~97%
Ramp-out week73%

The cohort held at the chain rate right up to the peak week, then lost a third of its on-time collections and only partly recovered the week after. The rest of the chain never left 97%.

Where · the collapse decomposed, one value per dimension

Curbside pickupfulfilment channel · share of impacted orders78%
Southeastregion38%
Beauty · Household Essentialsthe cross-department basket33%
Cleaning · Skincarecategory cluster29%

Contribution analysis of the on-time rate. It is one channel in one region — walk-in and ship-to-home in the same stores held. The department concentration is the giveaway that it is the 4-to-7pm collection window: a cross-department basket picked at peak.

Why this is the early warning and not the damage

Nothing in the sales line has moved yet — the week is a peak week and the totals look like one. Sentiment is the only surface where a fortnight of failed pickups is already visible, and it is visible now, while the labour mix for the rest of the holiday window can still be changed.

So what

This is the week the region can still be recovered rather than surveyed. Detractors were created by a specific, repeatable failure — a collection slot that was not met — in fourteen named stores, and there is still holiday window left to change the labour mix in them. Waiting for the sales line to confirm it means waiting until the guests have already decided.

The points that matter

Detractors were created, not converted

Promoters held at 40% while detractors rose to 30%. Nobody swapped sides on preference — a group of guests had one specific bad experience, in one specific slot.

30%detractors

The failure has an address

78% of the impacted orders are curbside, in one region, in a cross-department basket. That is fourteen stores and one collection window, not a regional sentiment shift.

78%one channel

The paid spend is amplifying it

Conversion-objective ads in Beauty and Household Essentials convert 26% below plan because they send guests into the loop that is failing. Every impression until the fix is a detractor bought at full price.

−26%attribution

Recommendation

Move the peak-window labour mix from the top-quartile remodel cohort into these fourteen Southeast stores for the remainder of the holiday period, and hold the conversion-objective spend in Beauty and Household Essentials there until collections are back on band.

Questions it already answers

Has this hit sales yet?

No, and that is the point. It is a peak week and the totals look like one. Sentiment is the only surface where a fortnight of failed collections is visible today.

Is it the region or the stores?

The stores. Fourteen bottom-quartile remodel locations in the Southeast fell to 62% on-time; the rest of the chain held at 97% through the same week.

What does the paid channel have to do with retention?

The conversion-objective campaigns in Beauty and Household Essentials are the ones sending guests into the collection slot that fails — so they are 26% below plan and, worse, they are the mechanism creating the detractors.