Retail Operations · Frontline AI

The work is 80% of the problem.
The model is the easy part.

Twenty-eight years leading large-format retail operations. I build the systems, and the operator capability, that make AI actually land on the floor.

28 yrs operations700-person storeFull P&L

DEFINED
ACTUAL

The Framework

Workflow-First AI™

Most retail AI pilots do not fail because the model is weak. They fail because no one owns the translation between the AI and the frontline workflow, and because the scoreboard used to judge that work was built for a pre-AI store.

Name the role, the Workflow-First AI Leader™, and rebuild the measurement, The Operator's Scoreboard™. That is the prerequisite for the J-curve to bend. Start with the work, not the model.

Try the Delta Auditor → Read the thesis →


The Field Guide

The Operator's Field Guide

The book behind the framework — twelve chapters on making AI land on the floor. Chapter 1 is free to read; the rest are available as early access. Pre-publication draft

The Case01–03
  1. 01

    The Pilot That Died on a Tuesday

    A six-month AI pilot dies in a single shift — and nobody touched the model.

    Read the chapter → · 18 min · free
  2. 02

    Why Most Retail AI Fails Before It Starts

    Four real explanations, none sufficient. The residual is the work no one owns.

  3. 03

    The Work Is the 80%

    AI’s first job isn’t to do the work. It’s to finally let you see it.

The Pillars04–06
  1. 04

    Start With the People Closest to the Work

    Diagnosis begins where the knowledge actually lives — on the floor.

    for operators
  2. 05

    Codify How You Lead

    Earn the right to see the real work, then make your method explicit.

    for operators
  3. 06

    Close the Translation Gap

    Intent dies in translation. The fix isn’t willpower; it’s artifacts.

    for operators
The Scoreboard07–08
  1. 07

    The Operator’s Scoreboard

    Re-rank what you measure, or miss AI landing and stalling entirely.

    for executives
  2. 08

    The Five Stages of Workflow-First AI

    Ad-hoc, Mapped, Validated, Codified, Compounding — and knowing which you are in.

In Practice09–12
  1. 09

    The Workflow-First AI Leader

    The role retail AI requires, defined — and why the org chart lacks it.

    for executives
  2. 10

    Responsible Deployment Is Not Optional

    The gate is the price of shipping. A real gate produces a documented no.

  3. 11

    The 90-Day Diagnostic

    A ninety-day sequence you can start Monday, without anyone’s permission.

    for operators
  4. 12

    Invert the 80/20

    Spend your attention on the unbuyable 80%. That is the whole method.

Early access

The rest of the Field Guide, as chapters are finished — for operators and the people building the capability.

One note per chapter release. No spam; unsubscribe anytime.

© 2026 Jordan K. Jones. All rights reserved. Pre-publication draft of Workflow-First: The Operator’s Field Guide (forthcoming).


Method in Practice

The delta is the signal

Composite case

The order the model did not see

An AI ordering system recommends the daily produce buy. The veteran orderer overrides it for a cold front and a stadium match nearby, and on those days she is right. The override is not an exception to suppress. It is the variable the forecast never had.

Composite case

The phantom out-of-stock

A shelf scanner flags gaps for replenishment. A third are a planogram change the system has not caught; the shelf is not empty. The scoreboard counts flags closed. The real job is the availability the partner's triage protects. Those are not the same number.


About

Operator-academic

Jordan K. Jones

I have spent twenty-eight years in large-format grocery retail operations, including leading a 700-person store operation with full profit-and-loss responsibility. I can cite Brynjolfsson, Nonaka, and Edmondson in one paragraph and name the produce-forecasting workflow that broke on a Tuesday in the next.

The frontline is where my work starts and where it is tested. Adoption is not coerced. Visibility is not weaponized. The people closest to the work keep their dignity.

Education
  • M.B.A. · St. Edward’s University2015–2017
  • B.S., Microbiology · The University of Texas at Austin1998–2001
Boards & advisory
  • Board Director · Imagine A Way2024 – present
  • Board Director · Alzheimer’s Association, Capital of Texas Chapter2023 – 2025
  • Vice President · St. David’s Advisory Board2022 – 2024

Contact

Whether you are building a retail AI capability or hiring someone to lead operations, the work is the same.