I'm based in Athens, and I'm the founder of UpCommerce Group, an e-commerce agency. We help Greek brands sell online, both at home and abroad, and we help brands from outside the country sell into the Greek market. Lately we've also started helping brands from the Balkans sell in Greece.

I came up through paid advertising. When I joined GrowthMentor three or four years ago I started out mentoring on paid ads, PPC, that side of things, and over time I've moved more toward e-commerce as a whole. That's really where the work lives now, the full picture of getting a store to sell, not just the ad account.

Every place does e-commerce in a different way, and two examples make it concrete. Here in Greece and across the Balkans, people don't want to pay before they choose cash on delivery, so they pay after they receive the product. That creates a real cashflow problem, because plenty of the time people just don't pick the package up and it turns straight back to you. That doesn't happen in the US or in Australia.

The second one is marketplaces. In Asia, something like 55 to 58 percent of buying happens on marketplaces. In Europe that number is much smaller. So the channels that carry your sales, the way you promote, the payment methods you need, all of it shifts the moment you cross a border. People assume selling into another country is their same store in a new language. It isn't. It's a different game with different rules, and the localization is the whole job: the payment methods, the local couriers, the last-mile delivery, the way people there actually decide to buy.

AI is everywhere in the conversation, but there aren't many real implementations yet. Where I expect the first big ones is data analysis and customer service, and after that, personalization. The promise is fully personalized websites built around the individual visitor. They say they can identify you not just by location but by your interests, maybe whether you're a man or a woman, even your size, in about seven clicks on the site. So if you normally look at medium, the category page quietly removes everything that's out of stock in medium before you ever see it. Agents rewriting the page to match what they think they know about you. It's a bit Black Mirror, honestly. Maybe they'll know more about our preferences than we do.

As for whether AI replaces the e-commerce consultant or the performance advertiser, I don't know yet. Right now it's a tool that genuinely helps us, and it isn't the biggest challenge we face. That's the honest answer.

Don't be afraid of it, but respect how different it is. E-commerce is challenging, and it works nothing like running a classic store, so at the start it's genuinely hard to understand everything at once. The people who do well are the ones who prepare, who learn how it actually works, who know their numbers and understand what's happening underneath. That's who gets the growth.

The part people get wrong is expecting it to be instant. You create something one day and you do not have revenue the next. You have to invest, and you have to be patient. So my advice is simple: don't go it blind. Take a consultant, talk through the real problems, and if you're selling into another country, talk through the localization properly, the payment methods, the local couriers, all of it. Do it correctly and the numbers come. The ones who treat it as a get-rich-overnight thing are the ones who get burned.