The coming months, Waag is working on MicroDonor. We investigate whether a system of microdonations would be a suitable business model for open-source developers, content makers and cultural organizations that prefer not to make money with the personal data of their users. Through this blogpost series, we keep you posted on our (re)search.
This post originally appeared on our website. Previously, we explained the background of web monetization, outlined our first technical findings and gave a general project update.
Public program
On Thursday May 18, Waag organized an evening program (to be rewatched - in Dutch - here) for experts and interested parties in the field of fair, open revenue models. MicroDonor: a revenue model according to public values ​​was the first kick-off of the project towards a wider audience, and thus the first opportunity to measure the value of a microdonation system outside our own Waag team. With more than 40 visitors and a lively discussion between experts from different backgrounds, we look back at an inspiring evening.
The event kicked off with an introduction to the Public Stack model, within which the MicroDonor project positions itself. Micro-payments for digital content and services offer perspective as part of the "foundation" of a public technology stack. Colleague Taco presented our research trajectory so far (also see the previous blog post). We then entered into a discussion with the invited experts on the basis of a number of statements, whereby the public was also encouraged to actively participate via the chat. The evening ended with a reflection by a researcher who has experience with revenue models for the creative and cultural sector on a European level.
The panel was a valuable mix of expertise: a software developer who works with XRP; a publisher at Follow the Money (FTM); the person responsible for IRMA from SIDN; an advocate for creatives from the Kunstenbond; and a CRM manager at the VPRO. The panel members all in some way had experience with the search for fair business models; whereas, for example, FTM and VPRO have in the past experimented with paid apps, Tikkies and a Blendle button, software developer Wietse actually set up a system of microdonations for Reddit and Twitter users through his XRP TipBot. In addition, Caspar offered a broader view of the creative sector from within the Kunstenbond. He highlighted that the cultural sector is increasingly facing digitization - only further reinforced by the current pandemic.
A number of themes were clearly at the forefront of the discussion. First of all, the potential of a system of microdonations was emphasized. It could be a fairer, more transparent revenue model, in which the user, instead of (unknowingly) paying with his or her personal data, is in control of what platforms or content is supported. As such, content and platforms that are now often offered online for free could gain renewed appreciation. In theory, the micropayment runs directly from the user to the provider (in line with Peer2Peer thinking). In addition, many organizations and content creators themselves, on the provider side, increasingly want to get rid of a system of advertisements - both for ethical and convenience reasons. It also offers more flexibility: at a time when users want to commit themselves less and less to one broadcaster, newspaper or other platform, microdonations offer a solution for making use of different media in an accessible way (for instance to read a couple of articles instead of a whole digital magazine).
On the other hand, challenges were also discussed. For example, the speakers agreed that a microdonation system must remain fair for large and small parties. Small parties like Follow the Money, who spend a lot of time and effort on quality articles that may not be read by a large audience, should not be penalized by low web monetization revenues when this model is based on as many "views" as possible. How do we prevent microdonations from becoming a perverse incentive that leads to ‘clickbait’ and only reinforces the ‘attention economy’? In addition, it is important that the income really does end up with the maker or designer, instead of the distributing party that is often in between. Perhaps it is necessary to involve an intermediary party, who takes on an infrastructural role (of the distribution of money) and, for example, draws up leveling agreements, currently often done in the cultural sector. Ideally, donations run directly from the user to the maker, but an intermediary - such as Coil - is also necessary to protect the privacy of individual internet users.
Overall, the event highlighted the potential added value of microdonations on existing systems. There is an increasing demand at both user and provider level for fair, transparent and low-threshold revenue models for online content and services, and web monetization could partly meet this demand. Whether it is realistic that it will completely replace other existing systems remains to be seen, but it is clear that microdonations are promising as one of several types of income streams.
Technical developments
We have also been busy behind the scenes. After the first 'naive implementation', in which we transferred fake money via a test network to our own 'wallet' while using the video conference service Jitsi through meet.waag.org (see also the first technology update here), we are now working towards a 0.1 version of the microdonor web extension. In our GitLab repository (viewable here for enthusiasts) we developed two components. One for the extension itself, and one for the "beneficiaries" registry. This second component is about the beneficiary organizations; how are monetized websites recognized by the web extension and what interested organizations receive a microdonation? So far, only Jitsi and Waag are in the register, but even within meet.waag.org there are more involved organizations. Ultimately, this database will be automatically supplemented via the code we write; for now, during the co-design sessions (on which more later) we will consider how to set up this register. Do organizations apply, or do we prescribe a set of conditions on the basis of which organizations are automatically "approved" for monetization? What party manages such a register?
The first component that we are developing is about the web extension itself. As explained in previous posts, this will be a Waag microdonation web extension that operates alongside the Coil extension and with which you as user can set the conditions for monetization yourself. In the current v0.1 this works as follows: The extension recognizes the platform you are using or the content you are viewing, and sends it to the server with the register of monetized organizations. In a next version we want to see whether we can load this database locally in the web extension. The component with the register returns a list of the beneficiaries (currently two), from which the extension then randomly chooses 1 to monetize - with the fixed rate of 0.36 cents per hour that Coil uses. In the settings in the drop-down menu of the extension, you can also choose which beneficiary you support, and you can set a maximum amount per recipient.
The Waag web extension then picks up the "receipts" (transferred amounts) from Coil and adds them up, so that you as user can read how much you have donated in total. This is stored on the user's local server (computer), so that privacy is not compromised as with storing data centrally. The micropayments are done in XRP, a cryptocurrency set up by Ripple. Unlike cryptocoins like Bitcoin, XRP does not run through a ‘proof-of-work’ consensus mechanism (requiring enormous amounts of energy) but instead through ‘trust-based validation’. The payments are therefore faster and more energy-efficient. The receiving party receives the microdonation in its virtual wallet in the form of XRP, after which it can convert it to another currency as desired.
Next steps
So far the microdonor browser extension v0.1. In the coming weeks we will further develop the back-end, after which we will also look for the desired design at the front. Furthermore, we will look more generally, beyond the web extension (which is linked to Coil and therefore remains dependent on the conditions created by them): what else can be done with the Web Monetization API? What do we recommend for the ‘ideal’ monetization provider, besides Coil?
We will further elaborate on questions from both the public program and the technical research trajectory in the co-design sessions on April 22, May 6 and May 20. During these sessions, we will brainstorm on web monetization with a diverse group of open-source developers, creators from the cultural and creative industries, and advocates of open technology and fair business models. More information about the co-design sessions can be found on our website.
Top comments (0)