Let me tell you about a task I used to dread.
We manage payroll across more than a dozen entities that operate on different schedules. Every pay period, someone had to log in to one portal, download reports, rename files, upload them to another system, and make sure everything matched correctly in QuickBooks.
If you treat it like a new employee and offer guidance, the quality improves dramatically.
It worked, but it was repetitive and time-consuming. My first instinct was the same as many people's today: I pointed AI at the problem. It helped with some of the manual work, but the process still took nearly an hour each week.
That wasn’t good enough, so I stepped back and approached the problem like a manager. I contacted our payroll provider, identified the real bottleneck, and built a process that automatically moved reports between systems. The result was a workflow that processed months of backlog in under 30 minutes.
Here’s the key point: I’m not a programmer. That’s the point of this article.
The biggest misconception about AI is that success depends on technical skills. In my experience, it depends far more on management skills.
Why Most AI Results Are Average
By now, most businesses have experimented with AI. The challenge isn’t adoption anymore. It’s about results.
Many people type a quick request into an AI tool, accept the first answer, and move on. The result is usually competent but generic.
Ask AI to write a product description, marketing email, or social media post without context, and it will produce something tailored to the average business. The tool has no idea who your customers are, how your store is positioned, what the adult retail environment looks like in your city, or what you would never put in print. Out of the box, it assumes a generic business and writes accordingly.
That’s why so many retailers feel underwhelmed. The problem usually isn’t the technology. It’s that the AI was never given enough information to do the job properly.
The real divide isn’t between stores that use AI and those that don’t. It’s between people who know how to guide and evaluate AI and those who accept the first answer they get.
Stop Treating AI Like a Vending Machine
The mental model that changed everything for me was simple. Most people treat AI like a vending machine. They put in a request and expect a finished product.
A better comparison is a new employee. A new hire may be smart and capable, but on day one they don’t know your customers, policies, or expectations. You train them, explain how the business works, and show them examples of good work while correcting mistakes.
AI works the same way. If you treat it like a stranger, you’ll get generic results. If you treat it like a new employee and offer guidance, the quality improves dramatically.
The good news for adult retailers is that onboarding employees is already part of the job. The skills needed to manage AI are often the same as those used to train staff, evaluate products, and oversee daily operations.
The Skills Retailers Already Have
When I look back on the payroll project, four things made it successful.
First, I understood the work. I knew how payroll flowed through the business, where problems occurred, and what a successful outcome looked like. AI couldn’t provide that knowledge. I had to bring it.
Second, I knew how to break the project into smaller steps. That’s not a technical skill. It’s the same process managers use to assign responsibilities to employees.
Third, I knew what “good” looked like. This is where many people stop too soon. The first version of my automation worked, but it wasn’t finished. It still required too much manual effort. AI handled the busywork, but the job still took an hour. I knew it wasn't done because I had a clear standard in my head of what “finished” actually meant. Recognizing that a polished, helpful-looking output is still not good enough is itself the skill.
Retailers do this every day. A buyer can sit through a compelling sales presentation and still know a product won’t work in their store. The same judgment applies to AI. Just because something sounds polished doesn’t mean it’s right.
Finally, I built on existing systems. Instead of starting from scratch, I expanded tools I had already created for other projects. That’s an important lesson. The second AI project is usually easier than the first because you’re building on what you’ve learned.
Getting Started Without Getting Overwhelmed
If you’re interested in using AI more effectively, don’t start with a grand strategy. Start with one frustrating task.
Pick the process you complain about most often. Maybe it’s writing product descriptions, managing inventory reports or organizing purchase orders.
Before you begin, define what success looks like. Treat the first AI response as a draft, not a final answer. Refine it. Ask questions. Keep improving the result until it meets your standards.
Most importantly, don’t be afraid to ask the tool to explain itself. AI can help you learn as you solve the problem.
There is a learning curve, but it’s much smaller than most people assume. The investment isn’t a computer science degree. What's required is the willingness to manage a capable, fast, slightly clueless new worker — and to hold them to your standard until the work is genuinely done.
The Real Opportunity
The retailers who get the most value from AI won’t necessarily be the most technical. They’ll be the operators who understand their businesses well enough to teach a machine how they operate.
AI doesn’t need a programmer standing over its shoulder. It needs a manager.
And if you’ve successfully trained employees, managed inventory, negotiated with vendors, or run a retail floor, you already have more of the required skills than you realize.
Zondre Watson is the general manager of technology and analytics for adult retail chain Ero-Tech. With a background in finance, chocolate and controlled chaos, he blends retail know-how with AI tools to keep 17,000 products moving smoothly.
