AI in CEE go-to-market:
what it accelerates, what it can't

Most LinkedIn posts about AI right now are written by AI. You can tell. Same rhythm in every post. Em-dashes everywhere. Fake-confident takes from people who've never run a P&L.

13 years running commercial operations across CEE. 8 markets. 4 industries. Country Manager, Commercial Director, GM, with full P&L. I've started markets from zero. I've taken over teams that were already broken. I've taken a FinTech licensing all the way through filing with the National Bank and PwC. I've rebuilt commercial engines that had stalled. I've made the hires that worked and the hires that didn't.

I've also spent the last 18 months using AI hard in this work. Not for content. For execution. So here's what I actually see. And what I tell CEOs who come to me asking for "AI-powered sales."

Where AI changed my work

I'll be honest. I was late to this. I started using ChatGPT seriously in 2024. Then I found Claude, and that changed things. Not "made my workflow nicer" changed. Different work, different speed, different output.

The change isn't where the hype is. The hype is on "AI sells for you," "AI books your meetings," "AI replaces your SDR." That's not what's happening. At least not at quality. What's happening is that the work that used to take weeks of analyst time, multiple tools, and constant supervision now takes days.

A real example. Romanian market entry, last year. The ICP work — figuring out who actually buys, why, with what budget, against which alternatives. Historically takes me 3-4 weeks. We did it in 8 days. Not sloppier. Better. Because we could enrich account data at a scale that wasn't economically possible before.

Outbound is where it's most visible. The personalisation that used to need an exhausted SDR writing his fortieth message of the day. That's done faster now. And often at higher quality. We've run 60-day pipeline builds in CEE markets where the outbound was sharper than what manual teams produced.

And pattern recognition. This one matters most. Recovery engagement earlier this year. AI picked up a drop-off cluster in the funnel. A specific deal stage where conversion fell off a cliff in one segment. I missed it in my own pipeline review. It wasn't smarter than me. It just had more attention to give. We rebuilt the deal motion around that finding in two weeks.

Where AI didn't change anything

Then there's the other half of the work. The half I spend most of my time on. The half that didn't get easier.

A Polish CEO doesn't want to be outbound-ed by a sequence. He wants to know who you are. Who you've worked for. What you're going to do for him. Whether you'll still be there when things go wrong. AI can't have this conversation for you. And senior buyers in CEE — anyone running a serious B2B deal of €100K+ — know exactly what an AI-generated message looks like. They discount you the moment they see one in a context that should have been personal.

Romanians have a phrase for one of these moments: "mă mai gândesc." Translates to "I'll think about it." Means "no, but politely." If you can't read which one a buyer means, no AI tool is going to help you.

Deal judgment is the same. Knowing when "no" is a real no. Or "not yet, give me cover for the decision." Knowing when to push, when to back off, when to walk. AI can summarise your call. It can't tell you what to do at the next meeting based on a buyer's micro-shift in tone three minutes in.

Team building. CEE sales coverage is not "find local people." Different talent pools. Different compensation expectations. Different cultural codes for selling. AI can find me candidates faster. It can't tell me when to hire local versus remote. Or which Bucharest-based hire will close deals in Prague and which won't.

Regulatory work is its own world. I've worked on a payment institution licensing in CEE, coordinating with the central bank, with Big 4 advisors, with internal product, with legal. The judgment about how aggressive to be on certain clauses, when to push the regulator, when to accept a slower path. That doesn't get faster with AI.

And there's one more thing. The hardest to put into words. Knowing whether your commercial engine actually works, beyond what the dashboard says. Pipelines look healthy until they don't. Forecasts hold, then collapse in the last week of the quarter. Reading what's actually happening in a commercial engine — that's pattern recognition you build by being there. Same region, same kind of buyers, over years. Hasn't been automated. I don't think it will be.

What I tell CEOs who ask for "AI-powered sales"

I tell them they thought right. And especially that they thought now, not in 2-3 years. Because in 2-3 years, the ones who waited will be losing or in trouble.

Then I tell them the honest part. AI won't run their business for them. AI won't replace their sales team. What AI does is take them to the next level. But only if there's someone who actually knows how to show them the way and build the infrastructure, the team, the discipline.

The companies getting this wrong treat AI as a replacement for senior commercial work. "We don't need a Country Manager, we have AI tools." That's how you end up 12 months later with an AI-augmented pipeline that converts at half the rate you were promised. And no one in the building who can read why.

Research, ICP, outbound, signal detection, pattern recognition — AI. Conversations, judgment, team, regulators, reading what's broken — operator. That's the model that works in CEE in 2026.