E-Commerce Search Is Broken

Try something for me.

Go to any major online store. Bol.com, Amazon, Coolblue, it doesn't matter. Type "laptop for school, max 800 euros" and see what happens.

You get 400 results. Sorted by... what, exactly? Relevance? Popularity? How much the manufacturer paid to be at the top? Nobody tells you. The first three results are sponsored. The filters are designed for people who already know what they want. And if you're just someone trying to buy a laptop for your kid, you're on your own.

This is the state of e-commerce in 2026.

The search bar that doesn't search

Here's what I find fascinating: we have AI that can write poetry, generate videos, build entire applications from a text prompt. And the search bar on the largest e-commerce platforms in the world still works like it's 2008.

You can't type a sentence. You can't describe your situation. You can't say "I need a quiet blender for a small kitchen, something under 80 euros." You type "blender" and you get 2,000 results. Price low to high. Price high to low. Customer rating. "Relevance." That's it. Those are your tools.

The platforms know this is broken. They've known for years. But they have zero incentive to fix it. Their business model is built on you scrolling. Every extra click is another opportunity to show you a sponsored product, a "frequently bought together" suggestion, a banner for something you didn't ask for. The confusion is the feature.

Who does this serve?

Not you. If you already know you want a "Philips Airfryer XL HD9270/90," you don't need search at all. You Google the model number and click "add to cart." The search bar is useless for experts.

And for beginners? The people who actually need help deciding? It's worse than useless. It's overwhelming. You read four review sites, each with a different "best of 2026" list, each suspiciously featuring the products with the highest affiliate commissions. You compare specs you don't understand. You spend an hour and end up buying whatever has the most reviews, hoping the crowd got it right.

The entire e-commerce search experience is optimized for the platform, not the person.

What if it just worked?

I've been thinking about this for a while. What would it look like if product search actually worked the way it should?

You'd describe what you need in plain language: "I want a good airfryer for a family of four, nothing too expensive." And you'd get back five products, curated by something that actually read the reviews, compared the specs, and understood your budget. Not 400 results. Five. With honest explanations of why each one is good, and what the downsides are.

No sponsored placements. No hidden rankings. Just: here's what's good, here's what's not, here's the best match for what you described.

So I built it.

BestePick

BestePick is a search layer on top of existing e-commerce catalogs. You type what you're looking for in your own words, and it gives you a curated shortlist with honest, independent advice.

It works by combining a product catalog API (currently bol.com, more coming) with an AI layer that does what you'd do if you had three hours and infinite patience: read all the specs, check the reviews, compare the options, and pick the five best matches for your specific situation.

Every recommendation comes with a "why this product" explanation and an honest downside. Because every product has a downside. If a review site doesn't mention one, they're not reviewing, they're advertising.

Is it perfect? No. The product data is only as good as the catalog. The AI occasionally misses nuance that a domain expert would catch. But it's already better than scrolling through 400 results hoping you'll stumble on the right one.

The bigger point

This isn't really about BestePick. It's about a pattern I keep seeing: tools that were built for the internet of 2010, still running in 2026, because the companies that run them have no reason to improve. Search engines that rank content by backlinks instead of quality. Social platforms that optimize for engagement instead of value. E-commerce platforms that optimize for ad revenue instead of helping you find what you need.

The technology to fix all of this exists. It's been here for at least two years. The incentive structures just haven't caught up.

Someone will build the better version of every one of these tools. The only question is whether the incumbents adapt first, or get replaced.

I know which way I'm betting.