From answering questions to doing the work: what a coding tool tells us about AI’s next move in retail

It’s Tuesday morning. You own five hardware stores. Before you’ve finished your coffee, you’ve reviewed stock levels at three locations, taken two vendor calls, texted a manager about weekend staffing, and started comparing prices on a bulk fastener order.

None of these decisions are hard. You’ve been doing this for 30 years. The volume is what kills you. Two hundred small decisions a day, and each one needs you — or at least, each one has always needed you.

Something just happened in the software world that’s about to change that math.

A coding tool did something new

Until recently, every AI tool worked the same way. You ask a question. It gives you an answer. You do the work.

Earlier this year, a company called Anthropic shipped something that broke that pattern. It’s called Claude Code, and it’s a tool for software developers. But what it does matters far beyond software.

Claude Code doesn’t just answer questions about code. It reads an entire codebase — thousands of files — then plans what needs to change, makes the edits, tests whether they work, and fixes its own mistakes. It does this for hours without a human touching the keyboard. One company reported their engineering teams working 2 to 10 times faster with it.

That’s not “AI-assisted work.” That’s AI doing the work.

The difference sounds subtle. It isn’t. It’s the difference between a calculator and an accountant.

The pattern that matters

If you’ve been in business long enough, you’ve seen every technology follow the same arc.

First, it answers questions. Search engines answered questions. Early AI chatbots answered questions. You ask, it tells, you act.

Then, it assists with tasks. Google Maps doesn’t just tell you where to go — it gives you turn-by-turn directions. AI started drafting emails, suggesting inventory reorders, flagging anomalies.

Then — and this is the jump that just happened — it executes independently. It doesn’t suggest a route. It drives the car.

Gartner predicts that 40% of enterprise applications will include task-specific AI agents by the end of this year, up from under 5% last year. Lowe’s has already deployed an AI assistant to all associates across 1,700+ stores. By 2028, Gartner expects AI agents to handle at least 15% of everyday business decisions autonomously.

The industry calls this “agentic AI.” In plain English: AI that does the work, not just talks about it.

Why this hits independent retailers differently

When Lowe’s deploys AI, they have a technology department, a budget, and a rollout plan. They have people whose entire job is figuring this out.

You have yourself and a few managers. You’re the buyer, the HR department, the pricing analyst, and the operations manager. You’re not ignoring AI because you’re behind — you’re ignoring it because you’re running a business.

Here’s the thing: that’s exactly why agentic AI matters more for you than for Lowe’s.

AI that answers questions is nice. AI that does the work is transformational for someone who’s already doing six jobs. Small business AI adoption surged 41% last year — jumping from 39% to 55% — and most of that growth is coming from owners who realized AI handles the tasks they never had staff for in the first place.

The economics are starting to show. A 2025 Capgemini survey of 1,500 executives found companies averaging a 1.7x return on AI investments — and 62% increased their AI budgets year over year. They’re not spending more because it’s trendy. They’re spending more because it’s working. For a five-store operation spending a few hundred dollars a month on AI tools, that kind of return is the equivalent of a reliable part-time employee — one who never calls in sick.

What “agent” looks like in a store

This isn’t science fiction. Here’s what retail AI agents will look like within the next couple of years:

An inventory agent checks stock across your five locations every night. It spots that store #3 is running low on PVC fittings and store #1 has excess. It evaluates whether to transfer between stores or reorder from a vendor — and it does it, within the rules you’ve set. You wake up to a summary, not a crisis.

A pricing agent monitors competitor pricing on your top 200 SKUs. When the big-box down the street drops their price on a key item, the agent adjusts your price within the margin range you’ve defined. You set the strategy. It handles the surveillance and execution.

A customer knowledge agent recognizes when a returning contractor walks in. It pulls their purchase history, knows they’re working on a framing project based on recent buys, and prepares a quote for the materials they’re likely to need. Your associate gets a head start instead of starting from scratch.

Research on multi-agent systems — where specialized agents work together — shows 45% faster resolution times and significantly more accurate outcomes compared to single-tool approaches. And companies that adopt early are seeing substantially higher returns than those that wait.

The reframe

The wrong question: “How do I compete with big-box technology budgets?”

The right question: “What decisions am I making 50 times a week that don’t actually need me?”

Claude Code proved something important: AI can now handle multi-step work that requires judgment, not just recall. It plans, executes, evaluates, and adjusts — the same loop you run through 200 times a day on vendor orders, staffing decisions, price checks, and inventory calls.

The gap between that capability existing in software and arriving in your store is closing fast. Not because retailers are building AI — but because the AI companies are building for retailers.

The store owners who thrive in the next five years won’t be the ones who become AI experts. They’ll be the ones who get honest about which of their 200 daily decisions are real judgment calls — the kind that require 30 years of hardware retail experience — and which ones just feel that way.

Back to Tuesday morning

Same owner. Same five stores. Same coffee.

The vendor calls, the stock levels, the staffing texts, the price comparisons. Which of those tasks needed three decades of expertise? And which ones just needed someone reliable to follow a clear process?

That distinction is about to become the most valuable thing an independent retailer can understand.

You can’t train 25,000 products. But you can make them searchable.

A customer walks in with a broken toilet flapper. They don’t know what it’s called—they describe it as “the rubber thing at the bottom of the tank.” Your new hire has been on the floor for two weeks.

What happens next?

At a big-box store, the answer is easy: point them to aisle 24 and wish them luck. But at an independent hardware store, customers expect more. They expect someone who can identify the part, recommend the right replacement, and maybe even explain how to install it.

That’s the competitive advantage of independent hardware stores. It’s also their biggest vulnerability.

The knowledge problem

Hardware stores carry around 25,000 products. A veteran associate doesn’t know all of them—but they know which plumbing fittings are interchangeable, what questions to ask when someone needs electrical wire, and which paint will actually stick to a deck.

This knowledge comes from years of answering customer questions, making mistakes, observing colleagues, and accumulating hands-on experience. One hardware store owner described it as “thousands of hours learning in classes and on the job.”

Thousands of hours. Not weeks. Not months. Years.

And here’s the problem: half of all hourly retail workers leave their job within the first 120 days. By the time someone starts getting good, they’re often already gone.

Why traditional training fails

The standard approach to new hire training looks something like this:

• A few days of orientation

• Shadow a veteran employee

• Learn the register system

• Figure out the rest on the job

In fact, 32% of retail employees report receiving no formal training at all. The assumption is they’ll figure it out on the job.

This works fine for jobs where the knowledge set is small. But you can’t teach 25,000 products. You can only hope new hires pick things up over time—and hope they stick around long enough to become useful.

Most don’t. Every employee who leaves costs you recruiting, hiring, and training all over again—typically 50-200% of their annual salary. For a store with 10 employees turning over half of them each year, that’s the equivalent of 2-5 full salaries spent just replacing people. And that doesn’t count the harder-to-measure cost: customers who leave frustrated because the new person on the floor couldn’t help them.

The experience gap

Big-box stores solved this problem by lowering expectations. Home Depot and Lowe’s are self-service by design. You grab what you need, maybe use a kiosk, and check out. If you need help, good luck finding someone who knows more than what’s on the label.

Independent hardware stores can’t compete on that model. Their value proposition is expertise—the helpful employee who asks the right questions and finds the right solution.

But expertise takes years to develop. And with 50% turnover in the first four months, most employees never get there.

What if knowledge was accessible on day one?

The equation changes when you stop thinking about training as knowledge transfer and start thinking about it as knowledge access.

A new hire doesn’t need to memorize 25,000 products. They need a way to find the right answer when a customer asks. They need to know what’s in stock, where it’s located, and what it’s used for—right now, on the sales floor.

That’s what AI assistants like FastQuery do. Instead of years of accumulated knowledge in someone’s head, you have instant access to product information, inventory, and problem-solving context.

As Eric Hassett, owner of Hassett Hardware, put it: “FastQuery puts all of the knowledge an associate would take thousands of hours learning in the palm of their hand on day one. A first-day associate can help customers like a veteran.”

The bottom line

Independent hardware stores compete on expertise. But expertise has traditionally required years to develop—time that most employees don’t stick around for.

The data backs this up: companies with strong onboarding programs improve new hire retention by 82%. But “onboarding” in hardware retail isn’t about paperwork and orientation videos—it’s about knowledge access.

The stores that figure out how to give new hires instant access to knowledge will have a structural advantage. Their employees will be confident on day one. Their customers will get the expert service they came for. And when someone inevitably leaves, the knowledge doesn’t walk out the door with them.

You can’t train 25,000 products. But you can make them searchable.

Store updates shouldn’t live on a whiteboard in the backroom

Hardware stores run on informal communication. A customer asks for something you don’t stock. An associate notices an empty hook. Someone spots water damage in aisle 3.

These observations matter. They’re the raw material for better inventory decisions, better customer service, better store operations.

But most of the time, they don’t get captured at all.

The problem with store communication

Here’s how it usually works: a manager does a store walk, jots things down on a piece of paper, then communicates each issue verbally to individual team members. If you’re lucky, there’s a whiteboard in the backroom with a running list.

The information lives in someone’s head, on a scrap of paper, or on a whiteboard that gets erased next week. There’s no way to ask “what issues came up this month?” and get an answer. No patterns. No history.

What if AI could read the feed?

We built a News Feed into FastQuery—a shared space where anyone can post store updates. Customer request, shelf issue, delivery delay, whatever.

The difference: FastQuery’s AI reads every post.

Ask “What’s going on in the store?” and you get a summary. Ask “What products have customers requested this month?” and the AI pulls together a list. Then you can create a task to follow up—without leaving the conversation.

It’s not magic. It’s just making unstructured information queryable.

See it in action:

Why this matters for independent retailers

Big chains have systems for this. They have corporate software that tracks everything, dedicated processes, regional managers who aggregate reports.

Independent retailers with multiple locations don’t have that infrastructure. The owner can’t be in every store. They rely on store managers to surface what’s happening—but the information gets fragmented across locations, stuck in separate whiteboards and verbal updates that never make it to HQ.

That’s exactly why this kind of tool matters more for them. One place to see what’s happening across all your stores. One less thing that falls through the cracks.