← The Automation Files
How I Work 6 min read · April 2026

I don’t sell features. I sell the why.

There’s a move I make in every client conversation that I learned in a phone store twenty years ago. It has nothing to do with technology. It has everything to do with people.

I was working at US Cellular when a woman came in — older, a little unsure, the kind of customer who made some reps quietly groan because the sale was going to take time. She wanted to see her grandkids more. That’s what she said. She just wanted to see them.

She wasn’t asking about a phone plan. She wasn’t asking about data rates or coverage maps. She was asking about her grandkids. And the answer had nothing to do with any of that — it was picture messaging, and someone patient enough to show her how it worked. That’s it. That’s the whole sale.

She came back. And she brought her friends.

The move has always been the same.

Same week, different customer. A businessman rushed in near closing, clearly irritated at having to be there at all. He needed a phone. He needed it to work. He didn’t have time for a demo.

I didn’t give him one. I handed him a Blackberry and told him exactly three things it would do for the way he specifically worked. He bought it in under ten minutes. That’s not because I was good at selling. It’s because I was paying attention.

I loved the tech. Genuinely. I knew every feature, every spec, every use case for every device on that floor. But I figured out early that knowing the tech wasn’t what got people to buy — or more importantly, what got them to actually use what they bought and come back when something else changed. What did that was finding the human reason underneath the question and connecting the technology directly to it.

“Find the human reason underneath what they’re asking for. Connect the technology to that. Be straight about what it costs and what it’s worth. Don’t promise what you can’t deliver, because you’re the one who has to show up and deliver it.”

That’s the move. That’s been the move for twenty years. The tools keep changing. The move doesn’t.

What this looks like in a web dev conversation.

When I moved into web development, I did it the same way. I didn’t know the technology yet — Michael did. My job was the client side. And what I knew how to do was listen until I understood what someone actually needed, then translate that into language a developer could build from.

Most client-developer relationships break down in that gap. The client says something like "I want it to feel modern" and the developer hears a design brief. The developer says "we’ll need to set up a custom module" and the client hears budget overrun. Neither of them is wrong. They’re just speaking different languages to each other and hoping it lands.

I live in that gap. It’s where the work actually happens.

Over time, as I learned more about how the technology actually worked, I got better at the translation — not because I became a developer, but because I understood enough to ask the right questions and catch the mismatch before it became a problem. That’s a different skill than building the thing. And it’s the one that keeps projects from going sideways.

And now in AI consulting.

The conversations I have now about AI sound exactly like the conversations I had in that phone store. Business owners know something needs to change. They’ve heard enough to know that AI can probably help. But they don’t know where to start, they’re worried about wasting money, and they don’t entirely trust the people trying to sell it to them — which, honestly, is fair. There’s a lot of noise right now.

My job is the same as it’s always been: figure out what you actually need, be straight about what it would take, and don’t promise anything I can’t deliver. I’m not going to hand you a 40-page strategy document and disappear. I’m the one who builds it and then has to answer your questions six months later when something changes. That keeps me honest in a way that pure consultants aren’t.

I don’t sell AI. I sell the outcome you’re actually after — the thing you’re tired of doing by hand, the question you can’t answer quickly enough, the process that costs you an hour every time it runs. AI is just the current best tool for getting there. Same as picture messaging was the right tool for that grandmother. Same as the Blackberry was the right tool for that businessman.

Why this matters when you’re choosing who to work with.

There are a lot of people who can build automations. There are a lot of people who can connect your Zoho to your HubSpot or build you a custom form. The technical skill is increasingly accessible.

What’s harder to find is someone who will take the time to understand your business before recommending anything — and then tell you honestly when the right answer is something simple, something you already own, or sometimes nothing at all right now. That last one is rare. Most people selling implementation don’t have an incentive to tell you you’re not ready yet.

I do. Because my reputation is worth more than any single project. And because the grandmother brought her friends.

Michelle Onizuka

Systems architect and co-founder of Onizuka Studio. Former paralegal, former telecom top performer, current AI and automation implementer. Has been translating tech into plain English for longer than she cares to admit. Read the full story →

More from The Automation Files

Behind the Build

I saw the wave coming. Again.

How We Work

You have the skills. I’ll sell them.

Tell me what you actually need.

Not what you think the technology should do. Just the thing you’re trying to fix. I’ll take it from there.

Start the conversation →