A case study on monetizing open-access content — Final Report

Project Update

As a publishing partner to thought leaders around the world, we’re consistently reminded of the importance of publishing openly. Where possible, we believe at Electric Book Works that well-researched content should be freely available to all who choose to access it. To maintain this ethos, the need for research into new, ethical revenue streams becomes even more important. As book production evolves and books are no longer only available in print but also as websites, ebooks, and PDFs, our understanding of how books bring in revenue must evolve too. We’re grateful to the Grant for the Web project for enabling us to experiment with web monetization as a means to generate income by conducting a case study on the use of web monetization on open-access content.

The first period of our project was spent working on the integration of web monetization support into a selection of existing open-content websites. This involved identifying and overcoming any technical constraints that arose during implementation, and coordinating the logistics of the experiment alongside our partner sites who agreed to participate in the research. We then ran the case study for a period of four months from 4 May 2021 to 4 September 2021.

Progress on objectives

Our project had two main objectives; collecting and sharing interesting data as a case study, and adding web monetization support to our open-source Electric Book template as core functionality. We believe that we have achieved both of these objectives.

Our research collection phase concluded in September, providing results that indicate only a very small fraction of organic visitors to our selected websites had web monetization enabled. These results are discussed in further detail later on in this report.

The web monetization functionality has been added to the Electric Book template, an open-source template that produces website, ebook, app, and print books from a single content source. It is packed with features for professional book production that benefit authors or publishing teams immensely when creating open-access content of a high quality. Users can now easily enable web monetization on any website created with this template by adjusting their settings and adding their pointer to the specified place.

Additionally, while adding web monetization to a site is simple, tracking its use is not since web monetization prioritises user privacy. Our experiment raised interesting technical and ethical discussions around tracking the use of web monetization and we feel that this is where the real value of our research emerged from. We will discuss this in more detail later on in this report.

Key activities

The primary activity for the case study was to enable web monetization on a range of open-content websites. Web monetization functionality was added to the following websites:

• bookdash.org
• core-econ.org/the-economy
• core-econ.org/espp
• core-econ.org/doing-economics
• bettercare.co.za/learn
• dailymaverick.co.za/gamechanger
• lawfulliving.co.za
• myconstitution.co.za

By implementing web monetization in the same way for the same period of time over several open-content sites, we are able to compare results across content genres and territories.

Note that we deliberately did not offer a quid-pro-quo for using a web-monetized browser, and we did not openly advertise that the site was web monetized. Our aim was to measure only whether there were users across a wide range of territories and genres that happened to be using web-monetized browsers, as a measure of the current prevalence of such users. We’ll provide those numbers below under ‘Results’.

Data tracking and collection

One of the main aims of the case study was to gather information about the number of visitors to each site who are using web monetization. Tracking this information in a robust and ethical way was a challenge, since the web monetization system is designed to allow users to support content creators without having to offer up any of their personal information or sacrifice any of their privacy, that is, not tracking users is built into the core principles and functionality of web monetization.

We considered several approaches and evaluated them on their ability to provide an accurate count of site visitors using web monetization, without compromising the integrity of the case study and that of the web monetization standard.

A number of technical limitations, alongside the ethical considerations, impacted the way in which we could monitor the use of web monetization:

• The non-linear cost function of Coil support prevented us from using dollars earned as a proxy for the number or duration of site visits.
• Traditional tracking services such as Google Analytics or Loggr were also unreliable as most common in-browser advertisement blockers and network privacy-protection tools will prevent information from being sent to these tracking services. We made the assumption that there would be a high likelihood that visitors who use web monetization would also use these privacy protection tools, which would lead to under-reporting of site visitors using web monetization from these tracking services.
• Lastly, the built-in dashboard on our domain hosting console was not sufficiently customisable for this kind of data collection – it could tell us about site visits, but nothing about visitors using web monetization.

Trial and error in the development of a robust, private, and ethical monitoring system for this experiment led us to settle on the following framework for data collection:

1. Monetization - external
   • Uphold: the Uphold wallet, with one Interledger Protocol (ILP) pointer per site in the experiment
2. Monetization - internal
   • HTML: the Coil monetization meta tag in the head of each page on each site, with the relevant ILP pointer
3. Tracking
   • JS: a JavaScript tracking script that uses the monetization API to detect when a visitor is using web monetization, and loads the tracking pixels
   • AWS: the AWS S3 bucket contain tracking pixels, and the associated metrics and tools that come with it

To elaborate, we opened one business Uphold wallet, and created one ILP payment pointer in US dollars per site in the experiment. Since we kept the currency constant across all sites, we could examine the amount earned with each pointer against the number of site visitors to verify the robustness of the results. This would also make it possible to accurately distribute any earnings to the respective site owners upon completion of the case study.

We then created an AWS S3 storage bucket, with a system of folders within it. One folder was created for each of the sites in the experiment, and within each of those folders, two subfolders. We placed identical 1px × 1px transparent pngs (68B) in each folder. We made the images as small as possible, so as to minimise any impact on the user experience or data cost to the user.

We wrote a small script, making use of the monetization API, to load these pixels into each site under certain conditions. Since AWS Cloudwatch can be used to count GET requests on items in S3 buckets, we could then count the number of site visit sessions, as well as the number of site visit sessions by users with monetization, over the course of the four month experiment.

By using this combination of the monetization API and AWS tools, we were able to count the number of site visits from users with web monetization active, without gathering any additional information about the user or their browsing session. We felt that this method allowed us to both maintain the privacy of the user and garner reliable information for our case study.

Results

Table 1: Total sessions compared to the number of sessions with web monetization enabled during the course of the case study (4 May 2021 to 4 September 2021).

While we used pixel tracking to measure sessions with web monetization enabled, we also have access to these sites’ Google Analytics. Google Analytics is not useful for measuring web monetization, but it does provide another view of overall site traffic that can be interesting. For example, some users’ browsers may block Google Analytics, some may prevent our pixel tracking, and some might do both.

Also, because of the way Google Analytics records sessions and page views, we cannot always get a clear picture from such data. So we present it here only for extra context.

Table 2: Total sessions and unique page views recorded on Google Analytics during the course of the case study (4 May 2021 to 4 September 2021).

* The Google Analytics for core-econ.org cannot provide sessions for a given ebook’s URL path. We only have a sessions count for all of core-econ.org, which we are not able to disclose in this report since it includes non-ebook pages that we are not responsible for.

Table 1 shows the results from the four month monitoring experiment, with the total number of visits to each site shown next to the number of visits by users with web monetization. As can be seen, only a very small fraction of visitors had enabled web monetization.

As we chose to gather such limited information about visitors, there is little we can hypothesize about the reasons for these results. The three websites that received visitors with monetization over the course of the experiment were southern Africa based sites, all spanning a range of content types. Book Dash is a “social impact publisher of free books for very young children”, Bettercare is a non-profit that provides free learning programmes to healthcare professionals and We Have a Game Changer is a commemorative book detailing the first decade of iconoclastic news service Daily Maverick. While Bettercare and Book Dash received the second and third most visits of our partner sites respectively, We Have a Game Changer received only 560 visits throughout the duration of the study, indicating that types of content and the users they attract should also be taken into account, alongside overall visitor numbers, when trying to increase the chance of monetized users visiting a site.

While we took precautions to avoid having our own team’s visits to those sites trigger a positive result, there is a very small possibility that some of those six visits were team members.

We were also constrained by our choice not to implement any quid-pro-quo for users. Apart from adding some generic disclosure text to their about pages, we chose deliberately not to promote web monetization on our partner sites for the duration of the study – we only wanted to see what proportion of organic visitors had web monetization enabled. Adding exclusive content for web monetized users, and advertising the availability of this content, would most likely see an improvement in the viability of web monetization as an income-generating mechanism in the short to medium term.

Another reason for the low number of visits could simply be the innovative nature of web monetization. It is still relatively new software that requires a certain level of interest and awareness of the work involved in creating free content and a heightened degree of tech literacy on the part of the user to set up their wallet. Over time, as the popularity of web monetization grows, these numbers will increase.

Overall, we believe that the most useful work to come out of this experiment is the technical research conducted to determine the best way to monitor the use of web monetization. We hope that this research can be shared with others who wish to conduct similar studies without compromising the privacy of users.

Communications and marketing

We published a blog post on our website outlining our case study and one of our partner sites, CORE Economics, also published a blog post about their participation in the study.

Once we have submitted this report and confirmed the Grant for the Web team is happy with it, we will also publish this report publicly.

What’s next?

Open-access publications, which provide content that is free to read online without any login, always face the challenge of staying sustainable while providing high-quality content for free. On the sites we publish, we’ve never sold advertising because we believe advertising is toxic to trust and have always advised our clients against this approach.

As a publishing partner to major organisations, we are excited to offer web monetization as an alternative to advertising. Including it as a core feature in our open-source Electric Book template means that any projects created with this template in the future will allow for the enabling of web monetization using this functionality in the HTML head, which can be turned on in the project settings. We will also reach out to clients who have created books with us in the past and let them know that this new functionality is available to them for inclusion in their projects.

Depending on the feedback we receive, we’re hopeful that in the future we can use web monetization on our client-based projects and potentially explore implementing an enhanced experience for monetized users while still maintaining our principles of publishing openly.

As an experimental project this case study was not without its challenges, particularly in determining the best way to monitor the use of web monetization, but we feel that overall it has been successful in pushing us to experiment with new technologies, examine our principles of publishing freely and expand our understanding of the ways in which modern book publishing can incorporate ethical revenue streams.

Relevant links/resources

Electric Book Works website
Electric Book Works twitter