Field-Tested Innovation · Facility Services

Creative Solutions: The Store Summary

How swapping a live dashboard for an email address fixed the one problem Tableau and Power BI couldn't: getting insight into a Business leader's hands when the store's own WiFi and BI servers can't be trusted.

The Problem

Dashboards were dying a slow, quiet death

Facility Services, like most of Walmart, had hundreds of dashboards. Most of them followed the same arc: a big launch, a burst of use, and then a slow fade as fewer and fewer people opened them. Nobody killed them on purpose — they just stopped being useful at the exact moment leadership needed them most.

Talking through it with several stakeholders — including Senior Vice President Hunter Hart — surfaced the actual root cause, and it had nothing to do with the dashboards' content. The dashboards were fine. The problem was where and when leaders needed them: while physically touring a store.

The Constraint Nobody Was Naming

Three failure points, all stacked in the same 10 minutes

A leader walking a sales floor to have an informed conversation with the store manager was fighting three things at once, and any one of them alone was enough to kill the interaction:

1. Bad in-store WiFi. Store networks are built for registers and scanners, not for a VP live-loading a BI dashboard on a phone.

2. Unpredictable BI servers. Tableau Server and Power BI Server uptime was inconsistent — and there was no way to know in advance whether the server would be up at the exact moment someone needed it.

3. Compounding failure. A slow network plus a shaky server doesn't average out to "somewhat slow" — it multiplies into "doesn't load," right as the leader is standing in an aisle with the store manager waiting.

The dashboard didn't need to be live in the store. It needed to already be in their pocket before they walked in the door.

The Fix

An email address, not another dashboard

The Store Summary Report sidesteps the live-connection problem entirely by removing the live connection. A leader emails a generic inbox with a store number in the subject line; a Power Automate flow does the rest, and a finished PDF lands back in the leader's inbox before they've even walked in the front door — no server session required once it arrives.

How it works

1
📧

Leader emails "Store Summary 540" — a plain subject line, nothing else required

2

Power Automate flow triggers on the new message in the shared inbox

3
🔍

Flow parses "540" out of the subject line

4
📊

Filters the master Power BI dataset down to just that store

5
📄

Renders the filtered view + open work orders as a PDF/Excel pair

6
📩

Emails both files back to the leader's own inbox

⏱ End-to-end: about 120 seconds

From there, the leader has a finished PDF sitting in their inbox — readable offline, on any phone, regardless of what the store's WiFi or the BI server happen to be doing at that moment.

Worked Example

What actually landed in the inbox — Store 540

Below is a recreation of the real report layout so the shape of it is concrete, not just described.

Note on the data shown below: the report layout, section names, and metric types are real. The specific values, names, phone numbers, and city have been fictionalized/altered for this write-up — the original PDF contains real employee contact information that has no business being in a shareable explainer.

Facility Services | Store Report

Store
540
Region
10
Sub-Region
10–19
Store Age
38 yrs
Delivery Model
Insourced
State
LA
Store Type
Supercenter
Store Manager
J. Alvarez
Regional Manager
T. Boone

Response Time

78%
Rank 5 of 31

Repair Time

100%
Rank 2 of 31

First Time Fix

71%
Rank 15 of 31
Aging (under 15d)
2
NPS
9
Open WOs
2
Active Projects
13
Cart Inventory
103.4%

Refrigeration Health — % time in target range

Energy Data — utility systems (kW)

FS Contacts (excerpt)

RoleContact
HVAC ManagerD. Whitfield
GM ManagerR. Castillo
Regional ManagerL. Fenwick
Companion Artifact

...plus the open work order list, oldest first

The PDF is paired with a second attachment: every open work order for that store, sorted with the longest-open items at the top — so the leader can pair "here's what the dashboard says" with "here's the specific thing that's been sitting for 3 days" in the same conversation.

Tracking #AgeTradeDescription
WO–4471982 3 days Pest Management Ants reported at a front-end register; inside/outside treatment requested (ants, beetles, spiders, moths).
WO–4482310 1 day FM – HVAC Rooftop HVAC unit not cooling over the sales floor; several other RTUs flagged not-cooling / comm-loss on the same call.
Companion Artifact

...plus a rollup of every active capital project

Store-level facility work rarely lives in one system. The report pulls a rollup across the three project pipelines that actually touch a store, so the leader isn't blindsided by a capital project mid-tour.

FM/PM Projects

2
TypeYearStatus
ST (Restripe)2022Active
ST (Restripe)2021Canceled

Smart Bundle Projects

2
TypeYearStatus
Smart Bundle2022Active
Smart Bundle2021Complete

Phoenix Projects

10
TypeYearStatus
HVAC Setpoint Adjustments2021Complete
HVAC Capital Upgrade2022Complete
Electrical Switchgear EOL2023Pending
Why This Beat The Dashboard

Same data, fundamentally different failure mode

Live Tableau / Power BI mobile appStore Summary Report
Needs a live server sessionYes — fails if the server happens to be downNo — PDF is fully rendered before it's sent
Needs strong in-store WiFiYes — store WiFi is often the weakest in the buildingNo — already in the inbox, readable offline
Getting to the right store's dataNavigate filters/prompts in a dashboard UISend one email with a store number in it
Training requiredYes — how to find, filter, and read itNo — it's an email attachment
What happens when the server hiccupsLeader gets nothing, mid-tourLeader already has the PDF regardless
Impact

What changed

~120 sec
From email sent to PDF delivered
Zero
Live server connections required in-store
Any leader
Self-service — no dashboard training needed
2-in-1
Strategic KPIs + oldest-first work order list, same inbox

The bigger shift wasn't technical — it was recognizing that "make the dashboard better" was the wrong problem to solve. The dashboard was never going to reliably load on a store's WiFi. The fix was to stop asking the store to serve the data live at all, and instead deliver a finished artifact ahead of time, in the one format that survives bad WiFi and shaky servers equally well: an email already sitting in someone's inbox.