Budapest Data Meetup Community Overview 2017

This blog post is a short version of the talk “State of the Union” by Mate Gulyas, which he presented at the Data Christmas Meetup at Budapest. Data Christmas is an annual event where all the data related meetups get together to reflect on the year and celebrate. This blog post focuses on the meetup community.

Method

We used the meetup.com API to crawl the data about the Hungarian meetup groups. We used the same API for crawling membership data. For the actual data, the crawler source code and the notebooks used for this analysis, please see the Sources section.

Meetups

The Hungarian data community is vast and diverse from the meetups point of view. There are 31 data related meetup groups with over 7500 unique members. The biggest meetup is the Data science meetup with over 2700 members. It is also the oldest, created in 2010, 7 years ago. If we analyze the membership counts for the different meetup groups, we can see a nice power distribution.

What about the quality of the meetups? The ratings can give us some idea about how the members perceive these meetup groups concerning value and quality. (Each meetup is a dot. X=member count, Y=average rating)

Even the lowest rating is over 4.5. That’s excellent. As we expect, the bigger the meetup, the lower the ratings are. The reason is that as the group grows, it turns into a more heterogeneous group with different seniority levels, diverse interests. When the topic is broad (think about the Big Data meetups that cover a lot of different technologies and topics), it gets almost impossible to satisfy everyone’s interests. Other factors might influence this as well, but the effect is clear.

The top 15 cities where members are coming shows us an intriguing insight:

  1. Budapest – 6441
  2. London – 89
  3. Aba – 46
  4. Berlin – 41
  5. Szeged – 38
  6. San Francisco – 31
  7. Budaörs – 28
  8. Amsterdam – 26
  9. Vienna – 24
  10. Székesfehérvár – 23
  11. New York – 20
  12. Debrecen – 18
  13. Zürich – 16
  14. Bratislava – 15
  15. München – 14

 

Budapest is the capital, all of the meetups take place here, so it’s not shocking that by far this is the first. The third, Aba is an interesting case. It’s a small town with a population of 4426. It’s surprising that it comes to third place between London and Berlin. The reason emerges if we take a look at the meetup.com registration user interface.


When selecting your hometown, Aba comes first in the list for Hungary. Aba is the choice for users who don’t want to specify the city.

The ratio of Hungarian and non-Hungarian cities is also unusual. In the top 15, if we don’t count Aba, only five are Hungarian. That’s just 36%. The rest, 64% of the cities, are from abroad. It aligns with the common belief that there is a massive brain drain (Human capital flight) from Hungary. There are a total of 372 unique cities where members are coming from.
Let’s have a look at how the memberships changed over time. We can’t plot all 31 groups as it would make our chart way too busy. Let’s concentrate on the four biggest meetups.

  1. Budapest Data Science Meetup
  2. Budapest Users of R Network
  3. Big-Data Meetup Budapest
  4. Big-Data Budapest

The membership counts follow the technology trends accurately. It is even more clear if we add the AI meetups to the chart.

In 2010 the Data Science Meetup was created. Starting in 2013, the Big data meetups are organized, and the memberships are growing fast. For the Big Data meetups to get to 1000 members, it took around 2 years. For the data science group, it took 5. It will be interesting to see how the new AI and deep learning meetup groups will evolve.

The pace of how quickly members joined was accelerating for years. 2017 is the first year when the five most popular meetup groups saw less new members than the previous year. The growth is still massive, but the acceleration seems to stop.

Presentation

Sources

The meetup data is created by crawling the meetup API. The data and the code are available on our Github.
The crawler is written in Python. For usage information, please refer to the README file. The dataset is a CSV file.
We released the membership data, the group data and we also created two graphs. For details please see the README file.

If you would like to receive more data analytics posts subscribe to our newsletter or follow us on Twitter.