I met Róbert Soós and the UH.ro team in the spring of 2023, just as they were launching their membership programme. We worked closely together for about a year and a half through MDIF’s Amplify Europe programme. They are based in Székelyudvarhely, a town of around 30,000 people in Romania, where the majority speaks Hungarian. It is a small market within a minority-language market, which is not usually where successful audience revenue programs are born. When we started, the idea that a local digital newsroom could generate substantial reader revenue there was nigh preposterous.
Robert is very interested in the business of getting people to pay for local journalism. Over time, you develop a feel for what will convert, which stories readers value and how to ask for money. Robert and his colleagues have developed that instinct, but they also want to test it: survey the audience, examine the cancellations, use AI to help analyse the data, find out where their assumptions are wrong. I know much larger newsrooms that would benefit from this level of attention.
By the time our formal work together ended, they had around 600 paying subscribers. Now they have more than 800. There are hard limits to what even a very good team can earn in a town this size, but what they built pays for actual journalism, in a place with few alternatives.
Recently, Robert sent me their latest analysis. I told him he should write it up. What follows was written in Hungarian, translated with AI and lightly edited by me.
My colleague Egyed Ufó Zoltán claims I beg for money in every newsletter I write. That’s not entirely true, but I’ll admit it: this piece really is about money. More precisely, about why a small local newsroom gets money from its readers, and why it doesn’t.
UH.ro is based in Székelyudvarhely, a town of about thirty thousand people in Transylvania, Romania, with a large Hungarian majority. There are five of us, and we are independent. Often of each other, too. We’ve run a membership model since May 2023. This September we had roughly 900 active supporters. Just over 800 of them were paying subscribers. About a hundred more readers had directed 3.5 percent of their personal income tax to us, which Romanian law allows (read more about the income tax redirection mechanism here). Reader revenue now covers a quarter of our budget. For a community this size that’s a decent result, but it isn’t stability.
Media Finance Monitor is where newsroom leaders, funders, and media entrepreneurs come to understand how money moves through the information ecosystem. We decode funding, business models, policy, and platform shifts and turn them into strategies you can actually use.
Paid subscribers get premium editions, full access to the archive and all our research reports, for about the cost of a pistachio éclair and a good espresso a month.
For deeper engagement, our Partnership Program offers one-on-one strategic advisory, curated briefings, and access to our internal databases and invite-only events.
In a small newsroom, everyone does everything. You write, interview, shoot video, send the newsletter, run Facebook, write grant applications, and try not to burn yourself out in the meantime. There’s rarely time to stop and look at what’s actually working, and why.
This autumn we stopped. We didn’t write a new campaign plan. Instead, we pulled together everything our own data could tell us:
the notifications from our subscription platform (Ghost) about cancellations and the reasons given;
article-level conversion data;
exports from Google Analytics and Facebook;
a reader survey that more than four hundred people answered, of which we analysed 234 responses, split into subscribers and non-subscribers.
Some of our beliefs held up, some others did not survive contact with reality.
1. They don’t pay for speed
In the survey we asked readers what matters most to them about UH.ro. 86 percent of subscribers named investigative work. Only 20.5 percent named fast news. Among non-subscribers, investigative work also came first (64 percent), just by a smaller margin.
The article data told the same story. Our best-converting article, not inflated by any campaign, brought in 3.7 subscribers per thousand readers. The weakest brought in 0.3. That’s more than a tenfold difference, and it had nothing to do with the headline or the format. The top of the list looked like this:
an in-depth interview with the managers of a major local employer about why they had just laid off sixty people;
a piece on how public money gets distributed;
an investigation into how the leaders of a local youth organisation spent public funds.
Our most-read stories, meanwhile, were about a drone alert and a house fire. They brought in more than ten thousand readers each, and not a single subscriber.
The common thread is the stakes. Readers want to know who decided, where the money went, who is responsible, and how it will affect their daily lives. Raw information is available elsewhere and is not particularly valuable, turning it into a story is.
News may still be important, it keeps you present, it builds trust, and it’s why readers come back tomorrow. It just has a different job. News tells you what happened. An investigation also tells you why, for whose benefit, and what comes next.
2. You don’t need a star reporter. You need beats people own.
Over the year and a half we looked at, we published 974 articles, which were linked to 961 paid conversions. 61% of those came from a single writer: my aforementioned colleague, the one who accuses me of begging for money. He wrote 12% of the articles.
At first glance, that’s the classic star-reporter model, and the obvious lesson would be: everyone should writer like him. But a closer look told us something else: he doesn’t live off one or two viral hits, his paywalled pieces brought in a median of four subscribers each, and his single best article accounted for 15 percent of his total conversions. Behind it is a body of work he has been building for years: local power, public money, consequences.
When other colleagues followed a story with similar tencaity, with their own information and a clear intent to hold someone to account, the subscribers showed up for them too. One of them, for example, followed the youth organisation case mentioned above, article after article.
Every regular reporter needs a story or beat of their own, with their name on it, that they follow over a long period. It could be a municipal investment, a public service, an institution or a local company. Over time they build up knowledge readers can’t get anywhere else.
It’s also an organisational question. It’s risky to have most of your revenue depend on one person’s work. Working together on stories can help other reporters learn from them. The aim is for more reporters to build expertise of their own.
3. Getting subscribers is easier than keeping them
Conversations about subscriptions usually focus on how many new subscribers you get. Our bigger challenge was somewhere else.
Of the subscriber groups that came in through specific articles, 54 to 90% are still active today. In December 2025 we ran a taster campaign: one month free, with card details required. Of its 220 participants, only 28% are paying today. To be precise:
40 percent cancelled during the free month, so they never paid at all;
of those who paid at least once, just under half stayed.
That doesn’t make the campaign useless. But you shouldn’t mistake the campaign’s gross number for lasting growth. For us, a campaign of two hundred people turns into roughly 55 to 65 long-term subscribers, not two hundred.
Monthly subscribers showed another pattern. 68 percent of cancelled subscriptions ended within the first two months, and about half of monthly subscribers are gone within roughly four months.
In other words, subscription work doesn’t end at checkout, that’s where it actually begins. In the first few weeks you have to show people what they got for their money: your best work, the stories you’re following, and a reason to come back next week. You need editorial attention, because retention is a content question too.
4. Half the people who left never meant to
This was the most uncomfortable finding, and also the best news.