Between the 21st of June and 22nd of September, according to European Forest Fire Information System (EFFIS) large wildfires burned about 148,000 hectares across Greece, Albania, North Macedonia, Montenegro, Serbia and Kosovo. That is 139% of the 2006–2025 average for the same weeks. However, the burden was not shared evenly. The five Western Balkan countries burned almost twice their usual area, while Greece had a quieter season than normal. That is why this report asks whether media attention followed the fires.

Last year, Pikasa tracked two weeks of wildfires in North Macedonia in "When Flames Spread, So Do Stories". This time we widen the lens to six countries and during the whole summer and place the fires next to the coverage. To do that, we combine Pikasa Analytics' media and social media data with data from the EFFIS, which tracks the area burned and the number of fires each week.

Key takeaways

• The fire was in the Balkans; the story was in Greece. The Western Balkans had 77% of the burned area but only 40% of the articles. Greece, in a below-average season, produced 60% of all coverage.
• Volume isn't resonance. North Macedonia wrote a third of the Western Balkan articles but drew a tenth of the engagement. Serbia wrote a fifth and drew almost half.
• Impact, not area, drives attention. Engagement peaked with deaths, threatened homes and calls for EU help, not with the hectares burned.
• The fires kept coming; the coverage didn't. Regional coverage peaked weeks before new large fires did. By the end of the season, each article drew about 25 engagements, down from about 80 at the peak.

Volume isn’t resonance

The first thing that stands out in the data is how completely Greece dominated the summer's wildfire coverage. Across all six countries, Greece produced 60% of all articles and its audience represented 69% of all engagement. Its outlets averaged 726 wildfire articles each, against 46 to 108 elsewhere, which points to rolling live coverage rather than more outlets.

Beyond Greece, though, how much was written and how much audiences responded followed different rules. The gap between output and response is clearest among the five Western Balkan countries.

North Macedonia wrote a third of the region's articles but drew only a tenth of its engagement, while Serbia wrote a fifth and drew slightly more than half of the audience's attention. Much of North Macedonia's volume reflects the size of its media market that covered the fires, or 218 outlets, which is twice as many as in any neighboring country. More outlets meant more articles, but not a bigger audience response. It is the same pattern Pikasa found in "Volume Isn't Reach".

To see where the audience actually was, the next step is to split the response between articles and social posts.

What drives attention is impact, not area

Adding the European Forest Fire Information System (EFFIS) data, which allows us to track the area devastated by the fires as well as the number of fires for each week, we can see that what actually moved audiences is not the size of a fire, but its human stakes. Attention peaked when fires threatened homes, took the lives of those fighting them, or pushed governments to call for outside help. Greece is the clearest example. Despite a below-average fire season, it drew 106 engagements per hectare for scorched area, four times in any other country in this report.

At the other end of the scale are the two most abnormal seasons. Montenegro burned almost three times more than previous years and Serbia burned almost seven times its usual area. Yet these two countries drew the least attention per hectare of burned area. Each country's engagement peak explains why.

Sometimes impact and area coincided, and sometimes they did not. In Greece, they coincided: the deadly week also accounted for 61% of the country's summer burned area. In Serbia, they did not. The Deliblato Sands fire was first detected on 21st of July and flared up repeatedly for a month, threatening villages near Bela Crkva. By the 26th of August, Copernicus put the burned area there and elsewhere at about 13,920 hectares.

Yet Serbian attention peaked only when the fire became a political story. The request for EU help came after opposition criticism that Belgrade had not asked sooner, and after a Russian Il-76 had already joined the effort. Serbian coverage tracked the number of fires more closely than the hectares burned.

Montenegro shows the other side of the same pattern. Its coastal fires near homes and beaches produced half of its summer engagement in a single week. But the fires that started between the 20th of August and 2nd of September account for 69% of its burned area, and by then weekly engagement had already fallen by more than 80%.

None of this is new - our earlier case study, “When Flames Spread, So Do Stories - Wildfires in North Macedonia in Media and Social Media” the most shared content showed flames approaching homes and exhausted firefighters. Audiences respond to people and homes at risk, not to hectares.

Where it burned vs where the story was

At the regional level, the mismatch is plain. The five Western Balkan countries had 77% of the burned area and 85% of new large fires, but only 40% of articles and 31% of engagement. Across the region, coverage bore little relation to where the fires actually were.

Within each country, however, the picture changes: week by week, coverage tracked local fires closely.

In five of the six countries, weekly articles rose and fell with local fire activity, and coverage never lagged behind the fires. North Macedonia shows the tightest link. Four straight weeks of 7 to 10 new large fires, from early August to early September, produced a matching plateau of coverage. Its weak engagement, noted earlier, is therefore an audience question, not a shortage of fires.

Albania shows how much timing matters. Its biggest fires came last. Its record week, 3–9 September, brought 22 new large fires and 7,385 hectares, yet drew only about half the articles of its mid-August peak.

Kosovo is the exception. Its coverage peaked in July, when EFFIS mapped no large fires there at all. Its engagement spike in the week of 7 September coincided with Albania's record week, which hints at cross-border attention among Albanian-speaking audiences.

The fires kept coming; the coverage didn't

Albania's late surge was not an isolated case. Across the Western Balkans, coverage peaked in the week of 10 August and engagement a week later. New large fires, however, kept rising until the week of 31 August.

That week, the region recorded its most new large fires of the summer yet had only half its peak coverage and a quarter of its peak engagement. Measured per new large fire, coverage fell from 300–550 articles in June and July to 60–70 by September.

This points to the fact that the audience may have tired even faster than newsrooms. Across all six countries, each article came with about 80 engagements in the four peak weeks (27th of July to 23rd of August). In the last four weeks (24th of August to 20th of September), that fell to about 25, below the early-summer level of 28. The drop was steepest in Greece (−75%) and Serbia (−71%), followed by Montenegro (−54%) and Albania (−34%). North Macedonia stayed flat at a low level, and Kosovo rose on its single September spike.

Splitting each country's fire season at its midpoint shows where the fade was strongest.

The fade is firmest in Albania and Montenegro, the two countries whose worst fires came last. Elsewhere it depends on exact timing, so the regional pattern is strong but not proven. Still, it is the same intensity without persistence that Pikasa measured in “When the Headlines Fade” coverage responds sharply to events, then moves on.

Methodology

Data collection covered articles from media outlets in North Macedonia, Kosovo, Albania, Montenegro, Serbia and Greece, along with posts on Facebook, Instagram, TikTok and YouTube. These were gathered with Pikasa's Topics Scanner using the wildfires topic definition. Weekly data from the EFFIS were then used to track the number of fires and the area burned in all six countries. The analysis covers the astronomical summer, from the 21st of June to 22nd of September.

Conclusion

This summer split the region in two. The Western Balkans carried most of the fire but only a minority of the story, while Greece, in a below-average season, dominated the coverage. This was not because newsrooms ignored their own fires: within each country, coverage tracked local fires week by week and never lagged behind them. What decided how much attention a fire received was its human impact, the deaths, the threatened homes and the calls for outside help, rather than the hectares it consumed.

That leaves a blind spot. Audiences tired faster than newsrooms, and newsrooms moved on before the fires did. Therefore, the late fires, like Montenegro's late-August blazes and Albania's record September week, were the ones most likely to be missed. For institutions communicating in a crisis, the lesson is to plan for the whole season, not just its first dramatic week. For anyone measuring attention, it is a reminder that volume is not resonance, and that the story audiences follow is rarely the fire on the map.