Hacking (Di)DIY online communities

by Giuseppe Catalfamo, Luca Colombo, Luca Mari: LIUC

DIY has changed and become increasingly complex in last few years. This is due to a wide spectrum of reasons that led us to describe it as a socio-technical phenomenon. We could address several isolated factors related to technology, economy, society, psychology, ... that, realistically, have broadened the range of individuals that decide to become DIYers, lowered the difficulty barrier to undertake DIY activities and, more in general, reshaped how we make things in 21^st century, but no single element seems to provide a really general explanation of what is happening around DIY.

Although in our DiDIY Project we are focusing on the digital part of the phenomenon, we only consider here one strong factor of diffusion, possibly the strongest, that has altered even the non-digital parts of the larger DIY context: the Internet and the new range of possibilities it enables.

Even before the Internet, groups of people united by a shared passion already existed. There were, and there are, clubs based on sports, hobbies and any kind of activities. The same thing happened for DIY, which originated local associations often based on specific branches of activities, like electronics, gardening, home improvement activities, cooking, and so on. But we cannot effectively understand today’s DIY looking at local communities only, many of them focused on specific matters and all limited to specific geographical areas.

With the Internet, people were finally allowed to share their interests while overcoming the geographical and communicational boundaries of the local, known, and confined networks of DIY enthusiasts. The Internet enables people to step over the old limitations and build international groups based on their passions. And it enables us to have “ready to use” information about DIY, in the context of communities that are virtual places where millions of individuals, pooled by their common DIY interest, share ideas openly, find inspirations, give and receive feedbacks, promote collaboration and help individuals and communities themselves grow. Knowledge becomes more and more open. Thanks to Web 2.0, everyone is now an expert amateur (Tanenbaum et al. 2013, p. 2604), able to publish contents and ideas, while, before it, just a few were enabled to do that.

Online communities have then incredibly broadened the boundaries of DIY, considering both the produced artefacts and the people involved. Step-by-step tutorials, high resolutions images and videos attached, critiques and feedbacks are a very strong contribution to the knowledge sharing and enhancing process. Even beginners can now understand every single detail of a project work and its complexities, becoming more capable and autonomous as makers.

This broadening effect is leading to hundreds of thousands of users joining online communities. Each user, after signing up, provides some sort of personal information and, if she/he wants to, can start publishing, i.e., sharing, her/his ideas and works as “projects”. Hence, these two entities (users and their artefacts) are a huge, global, and mostly untapped source of data to explore DIY and DiDIY.

(some papers previously published analysing data scraped from (Di)DIY websites are:
J. Moilanen, A. Daly, R. Lobato, D. Allen, Cultures of sharing in 3D printing: what can we learn from the licence choices of Thingiverse users? The Journal of Peer Production, http://peerproduction.net/issues/issue-6-disruption-and-the-law/peer-reviewed-articles/cultures-of-sharing-in-thingiverse-what-can-we-learn-from-the-licence-choices-of-thingiverse-users
S. Phillips, Shapeways vs. Thingiverse: An exclusive in-depth web traffic analysis, http://www.inside3dp.com/shapeways-vs-thingiverse-exclusive-depth-analysis
D. Mendis, D. Secchi, A legal and empirical study of 3D printing online platforms and an analysis of user behaviour, https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/421221/A_Legal_and_Empirical_Study_of_3D_Printing_Online_Platforms_and_an_Analysis_of_User_Behaviour_-_Study_I.pdf)

What about, then, planning and carrying out a thorough analysis of DIY and DiDIY in online communities? But how would we be able to discern relevant structures and emergent patterns from what could be nothing more than contingent behaviours of the online community we are observing?

We decided to consider multiple points of view (PoVs) for reducing possible bias, and to analyse in parallel both a community of DiDIYers and one of DIYers, so to observe the phenomenon both from within and from outside, of course under the supposition that DiDIY is a specific case of DIY.

In order to carry out this multiple PoV analysis, we identified the two currently most significant communities (here) as for users enrolled and projects published, that also satisfied our requirements:

• Instructables (www.instructables.com), which showcases artefacts from every possible DIY domain: we used it as a reference for the DIY phenomenon as a whole, hence for the “from outside” analysis. It counts today about 200,000 users and 170,000 projects;

• Thingiverse (www.thingiverse.com), which, instead, collects only 3D-printing-related projects and, thus, is perfect for a DiDIY-specific, “from inside”, analysis. Even with a narrower scope, it is bigger, with more than 400,000 projects and about 630,000 users.

While more extensive results of our analysis will be the subject of later publications, here below are illustrated the first, interesting results: the registration rate (Figure 1), i.e., the number of new users who signed up every month, and the publication rate (Figure 2), i.e., the rate at which new projects were published every month. For both Instructables and Thingiverse data start from October 2008, when the latter was opened (Instructables exists since March 2005).

Figure 1. Monthly registration rate of new users in the two analysed online communities (previously unpublished data).

Figure 2. Monthly publication rate of new projects in the two analysed online communities (previously unpublished data).

The first chart shows some significant differences between the two time series. The vertical axes have very different ranges, due to the different growth rate of the two websites’ users bases. Thingiverse’s growth multiplied by more than 20 times in the last 3 years and started to decline only in the last few months, keeping, however, a notable rhythm of 22,000 new users per month. Instructables, on the other hand, had a “boom” in 2013 and 2014, getting to a maximum of more than 5,500 new users in a month, after which it slowed back to 2,000-3,000 users/month. The only common structure of the two series seems to be the abrupt growth at the beginning of 2013. Their correlation, 0.49, reflects this limited correspondence, still a remarkable value for a social phenomenon.

The second chart conveys quite a different message. Again, Thingiverse shows growth rate remarkably higher than Instructables’s, with a maximum value of about 17,700 of the former versus the about 3,700 of the latter, but incredibly similar trends appear here, especially after the mentioned boom in 2013. The two patterns seem to be drawn together, and indeed their correlation beats an astonishing value of 0.98.

While more data should be taken into account to propose some specific and reliable interpretations, this first parallel observation, from outside and from inside, supports at least two of the basic hypotheses on which our Project was designed and is developing: first, DiDIY shares some common patterns with DIY, of which it can be considered a specific case; second, both DIY and DiDIY are spreading to new audiences and domains and this is happening, increasingly, in the form of DiDIY.