A New Way to Shop Is Finally Possible

Last month I wanted to get a gift for my friend Marcus. He's super into photography and just got into backcountry skiing. His birthday was coming up, and I had this idea — maybe a drone? Or a GoPro? Something that combines both hobbies somehow?

I went to a major electronics retailer. Typed "gift for photographer who also skis" into the search box.

You can probably guess what happened. I got cameras. Regular cameras. Nothing about skiing. Nothing about combining the two interests. Nothing that understood what I was actually trying to do.

So I did what everyone does. I opened the filters panel. Clicked "Cameras." Then "Action Cameras." Then realized drones are in a completely different category. Went back. Clicked "Drones." Now I'm looking at 200 drones with no idea which ones are good for someone who wants to capture skiing footage.

Eventually I just texted my friend Dave who's into this stuff. He immediately said "DJI Mini, lightweight enough to throw in a ski bag, great for that kind of thing." I searched for that exact model and bought it in two minutes.

Dave understood me. The website didn't.

We've all been here

This isn't a complaint about one website. It's just... how online shopping works right now.

You need running shoes for flat feet. The site shows 847 options and asks you to filter by "pronation control" — not a term normal humans use.

You want a couch that fits your awkward living room and survives your cat. Best you can do is filter by "width" and "material: fabric."

You're looking for a birthday gift for someone who's into cooking but already has everything. Good luck expressing that in dropdown menus.

What do we all actually do? We google it. We ask reddit. We text a friend. We give up and come back later. Or we just pick something and hope.

Here's the thing though — this isn't because the technology is bad. The recommendation engines and inventory systems behind these sites are often incredible. Millions of dollars of engineering.

But the front door — the search box, the landing page, the way you actually tell the site what you want — that part still feels like 2015.

Something changed

Here's what's different now.

Three years ago, "AI search" meant autocomplete. Maybe synonym matching. The technology to actually understand what you're asking for simply didn't exist at scale.

Now it does.

When someone asks ChatGPT for a gift recommendation, it asks clarifying questions. It understands context. It explains tradeoffs. That's not science fiction anymore — people do it every day.

Which means we can finally imagine something better. Not incrementally better filters. Not a chatbot bolted onto the sidebar. A fundamentally different way of connecting what you want with what exists.

What this could actually feel like

Imagine going to an electronics site and typing "gift for a friend who's into photography and backcountry skiing."

Instead of keyword matching, the system actually understands: this is a gift, it's for someone with two specific hobbies, and you're probably looking for something that bridges them or enhances one of them.

It might ask one question: "Is he more into capturing his skiing adventures, or is photography the main hobby and skiing is just something he does?"

You say: "He takes his camera everywhere on hikes, so I think he'd love to capture his ski trips too."

And then instead of 400 random products, you get three solid options:

Here's the tradeoff: the drone gets epic landscape shots but takes time to set up. The GoPro is grab-and-go but less cinematic. The Insta360 is the most unique but has a learning curve.

That's it. Three options. Explained. Tradeoffs clear. You can buy with confidence or ask follow-up questions.

The shift that's happening

This isn't about search being "broken" — it's about what's suddenly become possible.

For twenty years, e-commerce search was fundamentally a matching problem. You type words, the system finds products containing those words, and then you manually narrow down from there. That was genuinely the best we could do.

But now we can do something different. We can understand intent, not just keywords.

"Gift for a photographer who skis" contains a ton of information:

• It's for someone else (so my preferences don't matter)
• Two specific hobbies that might intersect
• Gift context means it should be interesting/delightful, not purely practical
• Probably looking for something that feels thoughtful

None of that fits in a filter dropdown. But it's exactly what a good friend or salesperson would pick up on immediately.

The technology to do this kind of understanding at scale just arrived. Which means every e-commerce experience we've gotten used to is suddenly... optional. We don't have to keep doing it the old way.

What changes

When shopping works like this, a few things shift:

This is where it's going

Every e-commerce company talks about conversion rates. A/B tests button colors. Optimizes checkout flows. Spends millions on marketing to get people to the site.

And yet the way you actually find what you want once you're there? Still basically the same as a decade ago.

That's starting to change. The technology gap closed. What used to require a human expert — understanding what someone really wants and translating that into specific product recommendations — is now something software can genuinely do.

The question isn't whether shopping will feel different in a few years. It's who figures it out first.

We're building the intent layer that makes this work. If you're curious about what's possible, we'd love to hear from you.