Affandy Johan | June 3, 2025

Urban Exodus During Mudik Triggered Shifts In Indonesia’s Mobile Experience

Eid al-Fitr, or Lebaran, which fell on March 31, 2025, is one of the most significant national holidays in Indonesia, marking the end of the fasting month of Ramadan. Beyond its religious importance, it is also the largest annual mass movement in the country, with tens of millions of people traveling from cities and towns to their hometowns in a practice known as “mudik”. Indonesia’s Ministry of Transportation estimated over 146.48 million people took part in mudik in 2025.

The mass movement from urban to rural areas during Eid creates temporary but rapid shifts in mobile network demand, challenging operator performance. This report analyzes changes in mobile experience across Indonesian regions and cities during the Eid period.

Key Takeaways

  • Mobile user activity shifted sharply during Eid, where Jakarta experienced a 30% drop in Speedtest® samples by the week of March 31, reflecting the urban outflow during mudik. In contrast, Central Java experienced nearly double the typical Speedtest volume, and East Java saw an increase of over one-third, as mobile users returned to hometowns for the festive period. 
  • Median download speeds rose in cities as congestion lowers but dropped in areas experiencing user surges. Jakarta recorded a median download speed of 57.94 Mbps during the week of March 31, a 20% increase from the previous week. In contrast, Central Java’s speed fell to 15.99 Mbps (down 38%), and East Java dropped to 25.91 Mbps, reflecting network strain.
  • Jakarta saw improvements with a page load time of 1.42 seconds and a video start success rate of 62.25%, following the outflow of residents during the festive period. Conversely, Central Java experienced increased traffic, leading to a slower page load time of 1.94 seconds and a lower video start success rate of 50.87%. These patterns highlight how changes in user distribution during the festive period influenced mobile experience across regions.

Eid travel pressure met with stronger network readiness from regulator and operators

To support mobile performance during mudik, Indonesia’s Ministry of Communication and Digital Affairs (Kominfo) and major mobile operators undertook coordinated efforts ahead of Eid. Kominfo deployed a monitoring force of 1,500 personnel and 29 vehicles across 386 key locations, including transportation hubs. Mobile operators also boosted capacity. XL Axiata, for example, doubled its network capacity and deployed hundreds of mobile base transceiver stations (BTS) to areas expected to see spikes in demand. Telkomsel, Indosat Ooredoo Hutchison, and Smartfren took similar steps, reinforcing coverage along highways, terminals, and popular travel routes.

Based on weekly aggregated Speedtest Intelligence® data of all technologies combined, from the week of March 17 to April 7 2025, reveals population shifts during the 2025 Eid al-Fitr period in Indonesia. This is indicated by a decrease in mobile Speedtest sample density in Jakarta and other major urban centers, coupled with an increase in provinces like Central Java and East Java. This reflects not just physical movement, but also shifts in network load and demand.

Animation of maps showing showing Weekly Speedtest Sample Size During Eid in All Regions of Indonesia | Weeks of March 17 – April 7, 2025

At the regional level, Jakarta experienced the most notable drop in sample density during the Eid week of March 31. Compared to the week of March 17, Jakarta’s sample density fell by about 30%, indicating a substantial outflow of users. By the week of April 7, sample numbers in the capital had rebounded by over 40%, suggesting that many users had returned shortly after the festive period. In contrast, Central Java and East Java recorded significant increases during the same period. Central Java saw its sample density nearly double between mid-March and the week of Eid, reflecting a large influx of people returning to their hometowns in the region. East Java followed a similar trajectory, with sample density rising by more than one-third. 

Outside Java, most provinces showed low and relatively stable sample density across the four weeks. Regions such as Maluku, West Papua, and North Kalimantan did not exhibit significant fluctuations, likely due to lower population density, limited mudik movement, or infrastructural factors. This contrast underscores the concentrated nature of mudik, where most of the population flow occurs within and around Java Island.

Chart of Weekly Speedtest Sample Size During Eid in Select Cities of Indonesia | Weeks of March 17 – April 7, 2025

At the city level, the trend in Jakarta was mirrored in other major urban centers. Surabaya, a major city in East Java, showed a similar pattern as Jakarta, with a 20% drop in samples during the Eid week of March 31, compared to the week of March 17. 

Mudik Migration Reshapes Mobile Network Loads, Boosting Urban Speeds and Straining Rural Connectivity

There was a clear relationship between sample density and mobile network performance data, which directly correlated with the population movements characteristic of the mudik period Major cities and urban areas with reduced user activity saw download speeds increase, while some regional areas with increased user concentration experienced performance degradations.

Jakarta, which recorded the highest drop in Speedtest samples during Eid week, showed a temporary boost in median download speed. The city’s speed rose from 48.36 Mbps in the week of March 24 to 57.94 Mbps during the week of March 31, a 20% improvement. However, by April 7, when users returned, speeds fell by over 27% compared to Eid week, returning closer to pre-Eid levels. Median upload speed also saw a similar pattern, increasing slightly from 15.76 Mbps to 17.5 Mbps, and dropping down to 15.92 Mbps during the same period.

The opposite trend is seen in Central Java and East Java, two regions that received a large influx of users during the Eid period. In Central Java, sample density nearly doubled by March 31, while download speeds dropped to 15.99 Mbps during the same week—down by more than 38% from the week of March 17. East Java followed a similar pattern. As sample density increased during the week of March 31, performance dipped to 25.91 Mbps, a decline from its pre-Eid baseline of 29.26 Mbps.

Mobile Performance Trend For All Technologies Combined In Select Indonesia Cities
Speedtest Intelligence® | Week of March 17th – April 7th

City-level performance during the Eid period mirrored the broader regional trends, with download and upload speeds rising in cities that saw lower user concentration and falling in those experiencing increased demand. Jakarta recorded the highest median download speed during Eid week at 57.94 Mbps, aligning with the drop in sample density size as residents traveled out of the capital. Surabaya and Denpasar also saw performance peaks during the festive period, reaching 47.57 Mbps and 51.8 Mbps, respectively, suggesting lower local congestion during Eid. Upload speeds followed a similar pattern. Cities with lower user load during Eid, such as Pekanbaru, Denpasar, and Yogyakarta, recorded the highest upload speeds, with Pekanbaru reaching 21.83 Mbps. Jakarta’s upload speed increased slightly to 17.5 Mbps during Eid, then dropped as users returned. 

Regional Performance Comparison Between Week of Eid 2024 vs. Eid 2025
Speedtest Intelligence® | Week of Apr 08, 2024 and Week of March 31, 2025

A comparison of Speedtest data during the Eid week in 2024 and 2025 shows a clear improvement in median download speeds across most provinces in Indonesia. Out of 34 regions, over 64% recorded faster speeds in 2025. Notable year-on-year gains were observed in Papua, which improved by 28.37 Mbps, Jakarta by 9.91 Mbps, and East Nusa Tenggara by 8.68 Mbps. These improvements reflect a stronger network response during the peak holiday period in 2025 compared to the previous year.

Users’ mobile web browsing and video experience tracked closely with regions’ download speed and network demand

Fast page load times and acceptable video start times are critical to the user experience, especially during festive periods like Eid, when users engage heavily with social media, share greetings, and stream video content. Speedtest regional data during the week of Eid (March 31) showed regions with sufficient network capacity and aren’t strained by user surges, experienced faster page load and smoother video streaming.

Users’ Mobile Web Browsing And Video Experience By Region
Speedtest Intelligence® | Week of March 31

Jakarta stood out with the fastest page load time at 1.42 seconds and the highest acceptable video start time rate at 62.25%, backed by the highest download speed of 57.94 Mbps. Other areas like Banten and West Java also performed well, with page load times below 1.6 seconds and video start rates above 57%. 

By contrast, provinces that saw increased sample density—such as Central Java and East Java—faced higher page load times at 1.94 and 1.65 seconds, and lower video start success at 50.87% and 56.1%. Remote and low-speed regions such as North Maluku, Papua, and Maluku consistently performed at the bottom. Page load times exceeded 2.1 seconds, and video start time success fell below 51%. These areas also had low sample volumes and limited network capacity, pointing to infrastructure constraints rather than temporary congestion.

Proactive measures and collaborative efforts ensured largely stable mobile network operations during the Eid period, effectively mitigating increased demand and minimizing disruptions to user experience during peak travel. We will continue to track mobile performance across Southeast Asian markets during significant festive periods. To find out more about Speedtest Intelligence® data and insights, please contact us here.

Ookla retains ownership of this article including all of the intellectual property rights, data, content graphs and analysis. This article may not be quoted, reproduced, distributed or published for any commercial purpose without prior consent. Members of the press and others using the findings in this article for non-commercial purposes are welcome to publicly share and link to report information with attribution to Ookla.