Project maintenance and conclusion

8.1 Long-term projects

After your project is set up and has launched, you will need to determine how to best maintain your project and support your volunteers over the lifetime of the project, as well as best practices for sunsetting or ‘ending’ your project. By asking volunteers to donate their time and effort to your project, you are committing to the effort of project maintenance and data sharing.

Some projects have a finite set of data that can be transcribed using crowdsourcing over a relatively brief period of weeks or months. However, if your project perpetuates for months or years using new data or even new Tasks and Workflows, you’ll need to consider how to sustain its community of volunteers and perhaps also financial support for administrative, technological, and intellectual labor.

Work with students (or other enthusiastic volunteers) when possible. I set up an internship for a group of students who helped me with the crowdsourcing project (and other documentary editing work). Their enthusiasm for the subject material was refreshing and enlivened my interest for what I already knew was a great project, but appreciated an injection of excitement into what I was doing. The students helped me select the best documents to work on for crowdsourcing. And while they were only around for a short time, I got a lot out of explaining crowdsourcing to them and thinking through the documents. - Serenity Sutherland, BCCCT Cohort Member

8.2 Maintaining volunteer communities

Communication (see Section 6) with current and potential volunteers is the foundation of long-term project maintenance. Regularly updating your volunteer community about the progress and current needs of your project is critical for continued engagement. Routine newsletters, social media posts, and Talk updates about what’s new and ongoing with the project helps volunteers to continue their transcription efforts. Moreover, press releases, media interviews, and other forms of broad outreach can help to engage new audiences particularly when it showcases progress made from initial volunteer contributions.

The usual output of Zooniverse projects is a peer-reviewed article, conference presentation or book. But there are many other possible research products and early results of your work can and should be communicated to your volunteers. Keep your volunteer community engaged by communicating what you, your team, or even volunteers have found that offers insight into your documents and the history they tell.

Project updates can be shared on Zooniverse via Talk and through project-specific newsletters (which are distinct from the Zooniverse-wide newsletter that announces new projects). These are emails that you can send to your registered volunteers after your project has launched. Newsletters should share your project's progress and will often bring people back to your project. To send out a newsletter, email a plain-text copy to contact@zooniverse.org and we will send it out for you.

8.3 Financial support for projects

Anyone is welcome to create and promote a project using the Zooniverse platform, but historically most of Zooniverse’s project creators are professional researchers who come from a university, cultural institution, or governmental organization. For those in such roles hoping to maintain a project for months or years, it can be crucial to secure long-term support for a project from their organization. Such support is helpful before a project launches.

8.3.1 Institutional support

There is a great deal of labor involved in building a Zooniverse project; monetary or in-kind support from your employer from the beginning helps move projects from ideas to launch. Since the Project Builder allows teams and individuals to prototype projects with minimal technical assistance, it is also possible to get a project up and running in order to show your institution a functional proof of concept. Consider discussing your vision for the project with your institution’s office of public engagement, marketing or communications department, or executive committee early on and keep them updated as your project progresses. The long-term funding needs of each project will vary significantly, but common contributions institutions make come in the form of assistance with external communications and design elements, administrative staff or research assistants, data processing and analysis staff and infrastructure, and dedicated workspace or even dedicated staff. Transcription projects tend to need more data processing and analysis support than other forms of crowdsourcing science and often seek help from information technology units in their institutions.

8.3.2 Grant and gift funding

While institutional buy-in always helps, it is rare for institutions to fund all the needs of long-term crowdsourcing transcription projects. Many researchers are expected to bring in external grants from governments or gifts from donors and foundations. Such external monies can accelerate the timeline of your project or sustain your project for a longer period than you or your organization could alone. The sources of grant funding vary by country, but an excellent primer for understanding the grant application process in fields where transcription projects tend to originate is Raphael Brewster Folsom’s How to Get Grant Money in the Humanities and Social Sciences. Grant application cycles can take over a year, so begin thinking about your project’s needs early and sketch out the aspects of the project that need the most labor or infrastructure in order for the project to succeed. Depending on your technical needs, you may want to consider reaching out to our team to collaborate on a grant application.

Obtaining gift funding is much more idiosyncratic than applying for grants, but the best place to begin is by speaking with someone in your organization who works in a ‘development’ or ‘foundation’ unit and guides relationships with existing and potential donors. You might be asked to draft a pitch of your needs and present that to them or even directly to donors. As with institutional support each project’s needs differ, but to help your project thrive in the long term, begin generating your applications and pitches by outlining your overall objectives and working backward to assess the costs associated with activities you’ll need to undertake to meet your goals for the project.

The transcriptions that your volunteers will produce are a significant output that must be made accessible to your volunteers and the public, generally. How teams share their transcription data varies, but you should strive to make it clear to anyone accessing your data what you did to generate it, how they can utilize it, and what you have done or plan to do to analyze it. There are many places available to post and preserve your data. If you are a member of a research institution, your library or archive likely hosts a data repository in which you can deposit your data and even your analysis of the transcriptions. There are alternative options that allow unaffiliated researchers to deposit their data, such as Dryad, and disciplinary-specific repositories such as OpenICPSR (social science) that enhance the likelihood that members of a specific field will engage with your data.

8.4.1 Recording your methods

Regardless of the repository you choose, you should work to provide strong contextual documentation about your project and the methods used to generate your transcriptions and find consensus among them. The best means for documenting the processes you utilized to create your project is to write a readme file. Readme files are text documents that accompany data and guide users through the who, what, when, where, why, and how the data was generated. Cornell University has put together an indispensable reference for drafting quality readme files that you should consult when you’re ready to write your own.

8.4.2 Packaging your data

In reductive terms, your project will generate some .json files. You could post those files in an open Google Drive and hope that interested parties now and well into the future know what they are and how to utilize and interpret them. But a much more effective means of data sharing is to package your data in a fashion that makes it findable, accessible, interoperable, and reusable (FAIR). This requires you to decide what to include in your data set (e.g., images, transcriptions, consensus transcriptions, etc.), document your dataset (in a readme, as in the step above), outline the reuse rights of your data (usually by selecting a Creative Commons license), and transform all files into non-proprietary formats (generic .txt files instead of Microsoft’s .doc file).

The University of Reading has produced a detailed guide to preparing your data for a repository, and the guide produced by ICPSR is also useful. As you consider which repository to store your data, prioritize those known to have clear preservation plans that will keep files in working order and publicly accessible for decades or longer. Accrediting entities like the CoreTrustSeal can help you to evaluate repositories you’re considering for your data set.

Additionally, you may want to consider publishing your dataset in a journal, such as the Journal of Open Humanities Data. The JOHD also provides a list of recommended repositories for their publication requirements.

Data repositories help to share and preserve your data for generations, but they act as places to preserve your evidence rather than your interpretations. The venues for making arguments from your transcriptions are countless and include books, articles, documentaries, exhibitions, courses, websites, apps, and many other media. Moreover, you can also amplify your interpretive products through traditional and social media channels. Regardless of where you release your interpretations, it’s critical to both acknowledge your volunteers’ labor and to make your volunteers aware of how their labor is leading to your findings. Sharing your interpretive progress with your volunteers during and even after your project is an important part of giving back to your volunteer community and can be accomplished through updates to Talk, newsletters, and other project news sources like social media or a project website.

8.5 Wrapping up your Zooniverse project

Though Zooniverse projects can take years and your analysis of their results can take even longer, there often comes a point when your project will need to conclude. This section will give you insights into this process, so that you know what to expect.

8.5.1 Sunsetting your project site

Active Zooniverse projects are listed at https://www.zooniverse.org/projects. When your project runs out of data, it will be automatically moved to the “Paused” category. When you are sure your project is finished, you must reach out to the Zooniverse team and ask us to move it to the “Finished” category.

While projects can be deleted, it is our goal for Zooniverse-approved projects to keep an active, public site available as long-term reference to the past effort. The idea is to provide a record of the project, its Talk message boards, and a long-term link that can point to the project's outputs that are available and accessible to volunteer contributors. You may want to consider leaving one or more of your Workflows set to ‘Active’ to allow future visitors to experience the method(s) you used to collect your project data.

When you feel your project has been entirely completed, thank your volunteers via an email newsletter and explain what happens next. Update your Results page accordingly, and don’t forget to submit project publications to our Publications page. We expect researchers to use the results of the project in peer-reviewed research, and to share results with the community. As detailed in the step above, Classification data must be made accessible within 2 years of the project’s completion, though sooner is preferable.

8.5.2 Preserving access to your activities

While your Zooniverse project site can stay ‘live’ indefinitely as a finished project, crowdsourcing projects and researchers’ attendant analyses often produce other research outputs–from books to websites to apps to exhibitions and beyond–that should be thoughtfully organized and made accessible to volunteers and the public, generally. The Endings Project, a digital archiving effort from the University of Victoria suggests that researchers with projects that include significant digital elements should consider three questions as their project concludes:

How should the project conclude?
How should the project’s dynamic features be preserved?
Where should the project be archived?

Each project must arrive at their own answers to those questions, but The Endings Project outlines a set of compliance principles to guide researchers through considerations that will help the digital elements of their work have a high likelihood of remaining accessible and even functional for decades. As you reflect on what you’ve accomplished through your Zooniverse project, imagine what you’d like someone to know about your process and results fifty years from now and work toward making those portions of your project sustainable to preserve and legible well into the future.