← All writingHome
Numbers17 September 2026 · 6 min read

Fifty-sixth in the world, and why the ranking is the least useful number

Serbia sits 56th in the global startup rankings, 14th in Eastern Europe, with 305 million dollars invested since 2017. Here is what those figures do and do not tell you about Belgrade.

Serbia ranks 56th globally in the 2025 StartupBlink rankings, down three places, and 14th in Eastern Europe. Between 2017 and 2025 the country attracted over 305 million dollars in startup investment. Funding jumped 705 percent between 2023 and 2024 while the number of deals fell 31 percent. Belgrade dominates the national ecosystem, followed by Novi Sad and Niš.

Those four sentences contain one genuinely important fact and three that will not change a single decision you make. The important one is the divergence between money and deals: capital rose sharply while the number of companies receiving it shrank. That is not broad growth, it is concentration. A handful of rounds moved the total, and the median company felt none of it.

Concentration matters because it changes what advice is true. In a broad market, the honest advice is to get in front of many investors. In a concentrated one, it is to be the specific kind of company the few active funds are already looking for, or to skip the round and build on revenue. Pretending otherwise wastes a founder's most expensive year.

The rankings themselves I would ignore. A position in the fifties, moving by three places, tells you nothing you can act on and quite a lot about how thin the underlying data is for a market this size. Nobody chose Belgrade because it was 56th and nobody left because it dropped to 59th.

What is actually distinctive here is density. In one city you can walk between the coworking clusters, the meetups, the alliance groups and the two or three organizations that run programs, in a single afternoon. Thirty minutes into a first meetup you will meet five people on exactly your path. That is not something rankings measure and it is the reason things get built here faster than the numbers suggest.

The other movement worth watching is state capital and infrastructure: commitments to AI infrastructure including a supercomputer, participation in European digital programs since 2025, an innovation fund with real money. Public money arrives slowly and with paperwork, and for the right kind of company it is still the cheapest capital available in this market.

So the number I would put on a wall is not 56 and not 305 million. It is the count of companies here that reached a first paying customer outside the region this year. That is the metric that eventually moves everything else, and it is the one nobody publishes.

If you are building in Belgrade and want a straight conversation about which of these numbers should affect your plan, write to me.

Reply

If any of this is wrong, or right in a way you can add to, I would rather hear it. Write to antanaskoviczarko@gmail.com or find me on LinkedIn.

Newsletter

Get new pieces by email

Occasional essays on AI adoption, regional cybersecurity, and building small products. No schedule, no marketing, unsubscribe whenever.

More writing

Numbers

Who actually writes the checks in Serbia

43.9 million dollars went into Serbian startups in 2025, and a small, nameable group of funds and angels wrote most of it. Here is the map.

Numbers

4.55 billion euro of exports, and the part that is not ours

Serbia exported 4.55 billion euro of ICT services in 2025. It is a real number and it hides the thing that actually matters for founders here.

Numbers

Seventy-two percent want to raise, and hiring plans keep missing

The Startup Scanner surveyed 170 startups in Serbia. More of them plan to raise than ever, and for the third year in a row they hired fewer people than they said they would.

Building without code

I am now a Lovable Certified Partner

What the certification actually means, why I went through it as a non-technical person, and how it changes the way I build products with founders and organizations here.

AI adoption

The gap is not the technology, it is the readiness

What I keep seeing between what frontier tools can already do and what organizations here are set up to absorb, and what actually closes that gap.

Regional cybersecurity

Nineteen startups in eleven countries

The mapping of early-stage cybersecurity startups across Southeast Europe, and why only two confirmed active product startups sit in Serbia.

Building without code

Non-technical, on purpose

Notes from shipping small products with AI tooling: what these tools genuinely do, where they stop, and why I would rather find out from the inside.

AI adoption

The 95 percent number, and what it actually says

MIT's NANDA researchers found that 95 percent of enterprise generative AI pilots produce no measurable return. Read closely, that is not a verdict on the technology.

Regional ecosystem

398 million euro, 109 deals, and what that means from Belgrade

Southeast Europe attracted 398.3 million euro of venture capital across 109 transactions in 2025, up nearly 40 percent. Here is what that scale actually implies for founders here.

Founders

How to build a startup from scratch

The honest sequence, written from the side of the table where I sit: what to do in the first ninety days, what to ignore, and where most people here actually lose the year.

Serbian ecosystem

One in three Serbian startups made no revenue at all

The 2026 Startup Scanner numbers are blunt: a third of startups here earned nothing last year, another third earned up to 50,000 euro, and a quarter are planning to raise over a million.

Serbian ecosystem

43 million dollars, and who actually got it

Serbian startups raised around 43 million dollars in 2025. Fewer deals, bigger cheques, and a selection filter worth understanding before you plan your own round.