By Shyam Rao, Founder and CEO
Every operator knows the people-process-tech framework. It has ordered change management for decades, and the order is the whole point: start with your people, redesign your process, and only then, at the end, choose the technology. Careful. Deliberate. Tech last.
AI breaks that sequence. Run it in the traditional order and you’ll spend 18 months aligning stakeholders around a technology landscape that changed three times while you deliberated.
We think the framework still holds, but it has to be reimagined for the AI era: the sequence reverses. Tech first, then process, then people. Here’s what each pillar looks like when you stop managing AI like a traditional technology risk, and start managing it like what it is: a massive accelerator.
Tech: incrementality is obsolescence
Traditional risk management treats a technology decision as something you get right once. Requirements, RFP, committee review, a 12-month selection cycle, a stage-gated pilot. Decide once, commit deep. For a fuel system or an ERP, that discipline makes sense. The choice is expensive to reverse.
AI inverts the math. The cost of waiting compounds every quarter, while the cost of an imperfect choice keeps shrinking, because reversibility beats perfection and you will swap tools as the market moves.
That flips where the risk actually sits: the safest-seeming decision, waiting for clarity, is the one that hands a two-year learning head start to the chain across the street.
So, pick something this quarter. Optimize for time-to-learn, for how fast your organization starts touching the technology. One CEO we work with framed his own decision in exactly those terms. For him, the ROI case was never the question, since the benefits of working with Tote are obvious. The question was whether he could afford to stand still.
Process: name the outcome, let early adopters pull the rest
The traditional rollout controls everything. All teams trained on launch day, progress held to the slowest team, champions assigned by title, success measured by who logged in. That process produces beautiful adoption dashboards and very little adoption.
Name one outcome instead, in units the business already trusts: cycle time, cost, time returned to the floor. Then pick a lighthouse workflow, something high-frequency with a clear owner and daily friction the field already complains about, where the before and after shows up in weeks.
And choose your pilot people as deliberately as your pilot use case. Champions live in the ranks, in the org chart’s blind spots: the tenured store manager everyone else calls before they call the district office. Stragglers follow peer proof, never mandates from above.
We’ve watched this work. At one multi-state operator, managers report reclaiming up to half an hour a day, with daily paperwork alone falling from 30 minutes to two. That matters in the daily life of a store manager, and multiplied across a couple hundred stores it becomes real operational capacity that can be deployed in higher-value activities like customer service and employee coaching.
The signal that mattered most was smaller: at 3 a.m., a new associate put her question to Genie AI, our assistant built into the point of sale, instead of waking the pilot manager at home, then told her about it the next day, proud she had handled it on her own. Wins like that recruit the next store faster than any mandate.
Measure the outcome you named, and skip the rollout theater.
People: “what happens to me?”
This pillar is last in the sequence – where adoption lives or dies. Traditional change management treats the people work as a training plan, a comms plan and a roadshow, assumes the questions are about features, and routes the concerns through HR.
The real questions are never about features, and what gets diagnosed as change fatigue rarely is.
Every employee in your company is quietly asking two things: Will I still have a job? Will my day get easier or harder? Fear doesn’t train away, and it doesn’t wait politely for the FAQ.
Answer both questions out loud, early, before anyone asks, and keep collecting signals weekly so you can adjust in days. This work belongs to the leadership team and can’t be tasked to HR alone.
Get this pillar right and the payoff shows up in a number every board cares about. Think about why frontline people actually quit. Standing in front of a customer without an answer is embarrassing. Apologizing for broken equipment, shift after shift, wears people down. Those small daily indignities push associates out the door long before wages do, and turnover is one of the largest controllable costs an operator has.
An AI rollout that puts answers and working tools in front of your people attacks that cost directly. One that ignores the “what happens to me?” question adds to it.
One motion, every function
These three pillars, Tech, Process, and People, when run in the new order, form a repeatable executive motion: set a small number of measurable outcomes, pick lighthouse workflows, establish guardrails, enable at scale, then measure and standardize. The same motion works everywhere AI touches your business: Pricing. Back office. Loss prevention. Merchandising. Finance and HR at headquarters.
The guardrails piece grows in importance as you expand. Your people already use AI on their phones every day. The choice is whether they get the internet’s answers or your company’s. Put your handbook, your procedures and your standards into the systems they touch, and the AI they were going to use anyway starts giving them your way of doing things.
Bolting AI onto a single silo repeats the mistake this industry has made with technology for decades. The tool was never the point. The point is an organization that absorbs new capability over and over, faster than its competitors can. The chains building that muscle now will spend the next decade compounding the advantage.
