Wikimedia Foundation Annual Plan/2024-2025/Goals/Infrastructure

The work we do at the Wikimedia Foundation has many facets, and has been described as supporting a "socio-technical" ecosystem. Within that ecosystem, the Foundation provides critical technical services, designs and launches new products, and innovates in areas like machine learning and internet-based collaboration. The work described in our "Infrastructure" goal outlines some of the key objectives and priorities in these areas for the coming year.

Advance Knowledge as a Service

Improve User Experience on the wikis, especially for established editors and functionaries. Strengthen metrics and reporting.

As outlined in this year's External Trends section, this work is informed by four key trends:

  • Search: Consumers are inundated with information, and want it aggregated by trusted people
  • Content: Contributors have many rewarding, potent ways to share knowledge online
  • Disinformation: Content veracity is more contested than ever before, and AI will be weaponized.
  • Regulation: compliance needs are growing and we need to increase global understanding of our community-based models

The importance of upgrading our technical infrastructure was also a recurring highlight in what we heard from volunteers in Talking: 2024.

Responding to these key themes – while also addressing the paradox that our projects are becoming both more vital to the knowledge infrastructure of the internet, while also becoming less visible to internet users – is the crux of the strategic challenge we face. In November 2023, the Foundation's Chief Product and Technology Officer posed a key question facing our Movement today: how do we ensure that Wikipedia and all Wikimedia projects are "multigenerational?" How do we ensure a sustainable future, and fulfill the "in perpetuity" part of our mission?

After months of feedback and discussion this year, we identified the following pillars to help answer these questions and guide our decisions and planning around future work:

  • Shape a changing internet.  Much has changed over the past 23 years. We need to bring encyclopedic content and volunteering to an internet driven by AI and rich experiences. This is our least well-defined pillar, but it's where we take our mission to the world and make it real. We are investing in experiments to realize the vision of being the essential infrastructure of the ecosystem of free knowledge online, and to shape the future through collaboration on research into how AI can support or hinder peer-to-peer production systems like Wikipedia.
  • Fund the future of ‘free'. We will sustainably fund our movement by building unified product and revenue strategies together. We encourage a culture of philanthropy at the Foundation, will support Wikimedia Enterprise as a partner to our API-focused work, and make targeted investments in donor-facing work on our websites and apps. We will continually look for additional ways to do our work more sustainably and efficiently.
  • Deliver trustworthy encyclopedic content above all. We need to enable the Internet's fundamental source of knowledge by addressing disinformation, misinformation and missing information in our ecosystem. We will continue our work on identifying knowledge gaps, and equipping volunteers with tools to fill them. And we will develop comprehensive approaches to scaled abuse and overcoming barriers to encyclopedic contributions across all languages. This requires clarity about MediaWiki's purpose and greater coordination between paid staff and volunteer developers going forward. We must also make tough decisions about what projects and features we support, and what we will stop supporting.
  • Fuel volunteer growth. This is the underpinning of everything we are doing as a Foundation. We will do that by nurturing multiple generations of volunteers. For both experienced and new volunteers, we must ask the question: "Why should I volunteer?" With that vital question in mind, we must align volunteer growth and opportunities to contribute. One recent example is "structured tasks," which has helped us draw in and retain new active editors. But much more is needed to become more sustainable and multigenerational in our work with contributors. This is the highest priority, and where we must begin our strategic shift.

Infrastructure: sustaining and future-looking investments edit

Our technical infrastructure is the underlying framework of hardware, software, networks, and services that support the operation of the Wikimedia sites and services. This is the base upon which our digital services, applications, and processes are built and operated. Included are a wide range of components including: hardware, software, networks, digital services such as hosting and security that support the operation, and the modernization and upkeep of our platform in an evolving landscape. To serve all these needs, we allocate time for both sustaining and future-looking investments. The sustaining work is primarily driven by ongoing maintenance needs of the unique systems that support the creation and distribution of knowledge, and the need to respond to incidents within these ever changing socio-technical systems. This work arises regularly throughout the year, has time permanently allocated for it, and is described in more detail below. The future-looking investments are informed by the external trends and our priorities as an organization. They are set out in the objectives and key results (OKRs) below. The work of the P&T department involves understanding the priorities in both areas, and how we shift resources and make tradeoffs to fix the most urgent issues, solve enduring challenges, and meet the needs of where we need to be.

A change we've made this year is more deeply incorporating these OKRs with the three other Foundation-wide goals. As we deepen collaboration across the Foundation, you will see the Product and Technology department OKRs showing up in the sections for other goals as well as Infrastructure.

Sustaining the technical infrastructure: work needed to maintain and support Wikimedia projects and volunteers edit

"The Foundation will make and keep useful information from its projects available on the internet free of charge, in perpetuity."

The Product and Technology teams dedicate a permanent, year-round priority to the maintenance, accessibility and operation of the wiki projects. This essential work starts with the fundamentals of developing and hosting a large popular website. We host our wiki projects in data centers, on servers and hardware we purchase, install and maintain, connected to each other and the rest of the Internet over a high speed network. We monitor and add capacity where needed, and refresh equipment when it gets too old. As an example, this year, we will be expanding our data center space in Carrollton, Texas, with some additional space and capacity for server equipment to keep up with growth of content and services on our projects.

We design and develop open-source software (most notably MediaWiki), and also use and deploy many existing third-party open-source applications, libraries and frameworks. Important bugs in our software get prioritized and fixed. Maintaining open source software requires highly skilled work from people with special expertise in open source software development, site reliability engineering (SRE), product management, program management, design, and more. Our staff work to ensure our software and systems are up to date and adapt to an ever-changing environment, including modernizing our code to continue to benefit from security fixes and to work well with new third-party software. For example, when software languages are updated, the old version becomes obsolete, meaning features don't work correctly. To avoid that we need to update our code to match. MediaWiki is written in PHP, and one of the things we need to do this year is to migrate from PHP 7.X to 8.3. Updating will make our systems more secure, load faster and be easier for staff and volunteers to work with them. This supports both a better experience for users and savings in terms of development time in the future thanks to security, performance and support improvements that come with language updates.

Over the past year, the unavailability of the Graph extension has been an ongoing technical challenge involving security and scalability considerations. This issue has affected readers and editors in many communities. In the coming year, we plan to build a new secure and scalable service to support the most common types of graphs used in the wikis, extensible for additional types of visualizations. We will work closely with community members as this plan becomes more defined.

To ensure our projects and content remain available on the Internet, in perpetuity, our teams dedicate a significant amount of effort on ensuring high-availability of our sites and services, and disaster recovery from catastrophic or malicious events. We ensure we have backups of important data, and are able to recover from them. Twice a year we test our ability to switch our sites between our data centers in automated fashion, and fix any issues we find.

Not all work is planned far in advance. We also respond to unexpected emerging events and incidents, like site outages, security reports or security incidents, or large scale vandalism attacks on our projects. We monitor our performance and barriers to reachability across the globe (including Internet connectivity problems, or censorship blocks), and investigate any anomalies we find. Some of these unexpected events or repeating patterns of problems result in staff prioritizing short-term follow-up projects that aim to mitigate or completely prevent further negative impact. As a concrete example: a pattern of availability incidents during high-profile celebrity deaths in the news used to overwhelm our wiki projects, infrastructure and services at all levels during the brief but very severe traffic spikes that followed. Through a combination of performance optimization, architectural redesign of bottlenecked areas, and capacity increases over several years, our projects and infrastructure have recently been able to withstand extreme, global traffic spikes (e.g. Queen Elizabeth's passing) without issues. The teams who do this work have knowledge about how to support billions of pageviews per month using open source technologies that are essential to the functioning of our socio-technical infrastructure.

Future-looking investments: objectives and key results edit

The Product and Technology teams also prioritize future-looking investment work. These are shared as "objectives and key results" (OKRs), whose impact tends to be framed as results that stakeholders or audiences can observe or experience directly. This is a continuation of the structure of work portfolios (called "buckets") that began last year.

We prioritized investments in three buckets: Wiki Experiences, Signals and Data Services, and Future Audiences. You can read more about why we chose these buckets in last year's annual plan. The Wiki Experiences work focuses on the experience of readers and contributors on the sites:  what they see and interact with on the wiki projects and in our apps. The work prioritized in the WE1 and WE3 below responds to the search and content external trends. WE2 responds to the content and disinformation trends, and is embedded in the Equity goal. WE4 responds to the disinformation trends and is embedded in the Safety and Integrity goal. Signals and Data Services focuses on insights into our content/metadata, making decisions on our content and services, and/or interacting with our content in structured or programmatic ways. Future Audiences focuses on experimenting with ways to invite new audiences into our Movement in direct and indirect ways and also aligns with the search and content external trends. You can find the specific objectives and key results for each below.

An "objective" is a high level direction that will shape the product and technology projects we take on for the next fiscal year. They're intentionally broad, represent the direction of our strategy and, importantly, what challenges we're proposing to prioritize among the many possible focus areas for the upcoming year. Each objective contains a few key results with measurable outcomes. Over the course of the year we describe multiple hypotheses that teams form for work, and we report quarterly on progress.

These objectives build off ideas we've been hearing from community members over the past several months through Talking:2024, on mailing lists and talk pages, and at community events about our product and technology strategy for the year ahead. You can view the full list of draft objectives below.

A note on budget planning: Across the Product & Tech department, we plan on allocating an estimated 50% of our resources to Wiki Experiences, 30% to Signals and Services, 5% to Future Audiences, and the remaining 15% to Product and Engineering Services.

These OKRs detail Product and Technology work that will happen in the service of the Infrastructure goal: