The Crowdfunder crawler was used to collect data for a case study aimed at testing the efficiency of the methods in identifying BTR organisations on Crowdfunder and exploring what insights can be drawn from the analysis of BTR activity data on crowfunding platforms.
The network visualisations were used in the evaluation of the methods and the histograms of the interactive map visualisation, in the analysis of the main trends in funding of the BTR projects found. The data used to build both visualisations, the networks and the interactive map, can be downloaded from the right hand side of this page.
The first step undertaken was to select a suitable seed project for the study. In doing so, the application was tested using different seeds, all of which had a social purpose and a large number of backers contributing also to other projects. From those seed projects, the one that returned the highest number of results leaving the keywords checkbox enabled – and using the default keywords shown in the input box - was selected. Namely, ‘Grow a Future for Families’, which returned a total of sixteen new projects presumably created by BTR organisations. The two files required to create the network visualisation with Gephi were exported from the app.
Then, the application was run a second time using the same seed. This time the keyword feature was unchecked, collecting a total of sixty-one projects run by BTR along with registered organisations1. The aim was to build a bigger network of projects in order to allow for a more accurate evaluation of the co-link analysis, as this method is being employed to track thematic relations between projects regardless of the legal status of the organisations behind them. This time, both, the files to create the network, and the statistic file to create the histograms linked to the map, were exported.
In order to create the two network visualisations, the corresponding files were imported into Gephi. Some of the settings in Gephi were adjusted to customise both networks for subsequent analysis. So for instance, nodes were resized based on number of connections to other nodes, and their colour was set based on the project category on the platform.
The later was done to allow for hypothesis testing: nodes sharing at least two connections or contributors are expected to have a similar theme, determined by their category on the platform, and therefore, be coloured alike. Then, the resulting networks were exported using the Sigma.js exporter, a plugin for Gephi that produces as output an interactive network in a web-enabled format2.
The second visualization type created comprises two different components: a map where the resulting projects from both platforms are placed – the sixty-one Crowdfunder projects and the thirteen Spacehive ones - and a histogram to be shown every time a project is clicked on the map. The data to create the map was manually collected from both platforms and entered into a single csv file. Each row in the file refers to a different project and contains the following information: the name of the project, its geographical coordinates -latitude and longitude-, its category on the platform, the name of the platform and so on.
For the sixty-one Crowdfunder projects, it is also specified whether the project is BTR or not. That is to say, whether it was found by applying the keywords filter or without using it. This file was then used to create the map with the Google Maps API. Likewise, the two statistics files exported from the app were used to create the histogram with the Google Charts API3.
Projects are located on the map based on their coordinates and represented by markers. The panel on the right side of the map allows to colour markers based on two different criteria: the platform where projects were found, and the amount of money they successfully raised. When clicking on a marker on the map, the histogram that shows the distribution of the money raised by the project appears below the map. Each column on the histogram represents the number of backers that has pledged a given amount of money.
In order to test the efficiency of the methods for Crowdfunder the following questions were formulated: do the resulting projects pursue a social purpose? Are these projects thematically related? To which extent platform categories can be used to determine that? With regard to those projects that are supposedly run by BTR organisations, is this the case? To answer the questions above, both network visualisations were used, although it was also necessary to dive deeper into the context of projects by looking up their pages on the platform and other external websites.
As to the first question, most of the projects on both networks were found to pursue a social purpose. The most common category on both networks is ‘Community’, and the second one, ‘Environment’, the category of the seed. While the number of projects is similar for both categories on the BTR projects network – with only one project falling under a different category -, on the network of mixed projects, the number of ‘Community’ projects is more than double the number of ‘Environment’ ones.
Moreover, projects from these two categories are highly connected through common contributors on this network. This might be due to the fact that these two categories inherently imply a social purpose, what is not the case for other categories on the platform. However, it seems to be another reason why specific projects from different categories are highly connected.
When going through the description of projects on the platform it became apparent that projects often do not stick to a single theme defined by their category, but rather, they present several thematic dimensions. As a consequence, it is usual to find highly connected projects that share a theme other than the one reflected on their category. In order to illustrate this, the different clusters of well-connected projects shown in the mixed network are presented as follows.
The biggest cluster, placed in the left side of the network, is made up of projects under the two categories already mentioned. As an example of the above, within this cluster there are two projects, ‘Snact’ and ‘The Happy Pig’, which reflect strong environmental values, although they were assigned categories different than ‘Environment’. And in fact, both projects, specially ‘The Happy Pig’, present a significant number of connections to projects under that category.
Likewise, the project ‘Vote for policies’ is highly connected to other projects with which it shares the theme of institutional politics. As this theme does not match an existing category on the platform, they were assigned other categories.
Another smaller but well-defined cluster appears in the bottom right side of the network. It is mostly made up of projects that match the category of ‘Sports’ and that have to do surf. However, not all surf-related projects in this cluster belong to that category. Two of them were instead assigned the category of ‘Community’. On the top part of the network there are two thematic areas, rather than differentiated clusters, loosely connected to the main cluster.
One of them is composed of projects within the category of ‘Food and drink’ or that otherwise revolve around a similar theme. One of these projects, ‘Made in Hackney’, which is about eco-community kitchens, is also well connected to projects on the other thematic area. Most of the projects in this area are about community spaces driven by environmental values and have been assigned the category of ‘Community’.
With regard to the last question, only six out of the sixteen projects on the BTR network were found to be run by BTR organisations. Namely: ‘Enchanted Acress’, ‘The happy pig’, ‘The Big Sun Flower project’, ‘A Tipi for Earth Education’, ‘Annan Harbour reopening’ and ‘Ask Amy’. In the next section, the context of these projects will be further analysed for a better understanding of the complex ecosystem in which BTR activity is produced.
The reason for that seems to be the inherent limitations of the method based on the keywords filter. Firstly, the fact that the keywords used to run the app were selected from the observation of a limited sample of projects. As a consequence, potential keywords that did not feature in the sample were neglected.
Another drawback of this method does not take into account the context of the keyword. Thus, all projects whose description contains any of the selected keywords will be discarded, regardless whether or not the keyword has being used to refer to the legal status of the organisation behind the project. Furthermore, it was found that among the results on the BTR network there are instances of projects run by registered organisations that only revealed their legal status on their own websites, but not in Crowdfunder.
As a conclusion, the assumption upon which co-link analysis is based proved certain to a great extent, as most the projects returned by the app have a social purpose and are often thematically related to those other projects with which they share connections, although not necessarily through their category on the platform. Therefore, co-link analysis seems to be an appropriate method to map thematically related activity on Crowdfunder. However, the analysis of the results on the BTR network showed evidence that the keywords filter should be further developed or substituted by other methods.
In this section, the specific context of the six BTR projects identified will be further analysed. In doing so, I will look, among other things, at the specific ways in which the organisations behind them are supported by other organisations in order to allow for the effective development or /and continuity of those projects.
Both ‘Enchanted Acres’ and ‘A tipi for earth education’ belong to the same category than the seed, ‘Environment’. The former project is defined on the platform as a ‘permaculture and skill sharing project’ and was created by a couple based both in Portugal and Cornwall. The project intends to provide a space where people can learn about sustainable ways of living. To that aim, they will offer people accommodation and workshops at no cost.
The money raised by their crowdfunding campaign will be employed to improve the facilities of their new space in Portugal. In order to ensure the continuity of the project, they plan to build and sell tipis and offer the site as a holiday destination, commercial activities that are likely to be somehow registered. Therefore, the space facilities will be used complementarily for business and BTR activities.
Moreover, this project is part of the Crowdfund Cornwall Campaign on the platform, aimed at raising money for projects on that area. The campaign has established a partnership with the Cornwall Community Foundation, which has pledged a total amount of £50,000 to fund projects within the campaign.
‘A tipi for earth education’, is a project aimed at raising funds to build a tipi for school visits where children can learn environmental values and connect with nature. The project was created by a couple, who in their spare time set up the Cambridge Sustainability Centre on their organic farm in Cambridge, where they run courses to teach people how to live more sustainably. The schools visits are part of its activity programme. In order to make the centre financially independent, they intend to offer weekends stays in the tipi. Once again, economic and BTR activity will be carried out in the same facilities.
The next four projects were assigned the platform category of ‘Community’ on the platform. Despite of that, the project ‘The happy pig’ presents a lot of similarities with the above projects. This project is aimed at converting a pig barn into a space fitted with dorms and learning resources for people to explore a variety of topics, lots of them similar to the ones mentioned on the Enchanted Acres project.
The project, which intends to provide all of that free of charge, is being created by a group of individuals at ‘An Teach Saort’, their permaculture and gift economy based small holding in Cornwall - although evidence has been found that is actually based on Ireland. I could not figure out about whether they plan to carry out some sort of commercial activity in the holding in order to make their BTR activity sustainable in the long term, although it might well be the case.
Behind the ‘Big Sunflower project’ there is a person living with the rare neuromuscular condition centronronclear myopathy. She set up the Information Point for the Centronuclear and Myotubular Myopathy website in order to raise awareness of the condition. With the same aim, she has also been running the Big Sunflower project once a year since 2011, consisting of sending out sunflower seeds at no cost to be grown by participants. Her project on Crowdfunder intended to raise funds to buy the stamps and envelopes for the last year campaign.
As opposed to the previous projects, this is a relatively small one, run by a single person with virtually no resources nor spaces where to develop her activity. The fact that this small project is not intended to grow into a bigger venture might explain why in this case BTR activity does not appear intertwined with commercial activity nor is supported by other organisations.
‘Annan Harbour reopening’ was created by a group of local volunteers with the goal of bringing together the necessary resources to clear the local harbour in the town of Annan of the silt accumulated over 50 years and turn it into a community asset. The group has been working since 2011 in a number of initiatives to improve the harbour facilities. At the time the project was created on the platform, it had been granted funding by the European Fisheries Fund (EFF) under the condition of getting a complementary source of funding to ensure its viability.
‘Ask Amy’ is an app prototype aimed at increasing the number of people in their 20s and 30s accessing and getting involved in politics in order to improve the legitimacy of democracy. The project on the platform intended to raise money to develop the prototype into a fully functional application. The organisation behind the project, No One Ever Told Me About Politics, is run by a group of volunteers coming from a range of private, public and third sector organisations. They have been sponsored by the Firefly Communications Group and, at present, are looking for other sponsors to support their job.
The amount set as funding target for the projects above depends on their specific aim and scope. This explains why this amount is that much lower for the ‘Big Sunflower project’: While the funding target of the other five projects was above £2000, this small project only aimed at raising £250.
Moreover, the higher the amount set as funding target, the higher the number of backers, although the correlation between the two variables is not very strong. Another trend that can be observed on the histogram displayed for each project on the map visualisation is the negative correlation between the number of backers and the amount pledged. For all projects, most of the backers have pledge an amount in between £50 and £100 and the number of backers gradually decreases for the next money brackets.However, quite often histograms show a number of outliers that have pledged high amounts of money rather far from the average. These backers are likely to be potential participants in the projects and/or benefit in some way from their realisation. Thus, for instance, the main backer of the project ‘Annan Harbour reopening’ is the project owner himself.
1Data collection from both platforms was carried out on 25/11/2014.
2When opened with a browser, the whole network is shown, and on its left side, there is a panel with information about the network and features to navigate it. The feature right on the bottom allows to filter projects by category. When clicking on top of a node, only the nodes connected to that specific node will remain visible and a right side panel will pop up with information about the project: its name, URL and category on the platform and its number of connections to other projects, which are listed right below, together with the username of the backers that contributed to both projects.
3The first step in creating the visualisation was to merge the statistics files from both platforms into a single csv file. Then, this file together with the file for the map were uploaded into a database and stored into two different tables. Data in both tables can be matched through their common field ‘project id’. Thus, when a marker on the map is clicked, the histogram gets fed with the specific statistic data for that project.