#10 Why Wikipedia doesn't do ads?

Read Time: 12 minutes

Aug 27, 2020

Author's Note
Welcome to the ~1,500 subscribers who joined after Vivek's piece on Jio. If you want to read more of our older pieces, you should definitely check out this piece on venture capital and this one on Dunzo. On to a bunch of updates.

We are launching a referral program! If you like our work, refer your friends here to rack up a high score and win some exciting stuff (not to be revealed now :P). We are going to be tracking this through a soon-to-be-launched public leaderboard. Shoutout to Bharath and Sainath at Nodal for setting up our referral program in 5 minutes. 5 minutes!!!!

We got featured in Substack's newsletter yesterday. We are one of the 3 newsletters from India that were highlighted in the post. Yayyy!

On to today's piece...

“Imagine a world in which every single person on the planet is given free access to the sum of all human knowledge”

This is what Jimmy Wales says when asked what's Wikipedia doing. Wales, Wikipedia's founder, is not claiming tall things. Wikipedia is the most sought after destination online to learn about anything under the sun. It has over 54M articles published in 285 languages to date. It receives close to 1.5B unique visitors per month making it the 14th most visited website on Earth ahead of Amazon, Netflix and Instagram.

If you are one of those 1.5B visitors and are based out of India, you'd have seen the following message on almost every Wiki article you visited in the last month or two.

All of us knew Wikipedia offers its content for free but not many of us wondered how they made money. Thanks to this, many readers (at least in India) now know that Wikipedia is run majorly on donations from both individuals and corporations. Wikipedia, in fact, is owned by the Wikimedia Foundation which is a non-profit organization. And they run fundraising campaigns like this every year.

Interestingly though, Wikipedia never planned on taking your donations when it started. They actually wanted to earn through ads. Why did they not go down that path? The answer lies in Wiki's origin story.

Origins of Wikipedia

Jimmy Wales took up the job of an options trader after completing his PhD in Finance in 1994. This was around the same time internet startups were springing up. Realizing the growth potential that the internet promised, Wales left his job to start Bomis in 1996. It was a website for erotic photographs and soft pornography offered via paid subscriptions for users. Due to the internet boom in the late 90s, advertisers paid good money for the ad space on Bomis.

It was while running Bomis that Wales began to experiment with a side project called Nupedia, sometime around 2000. Nupedia was a free online encyclopedia for scholarly peer-reviewed articles on a variety of topics, a reflection of Wales's love for the Britannica Encyclopedia during his childhood days. It was launched as a for-profit venture under the Bomis group with a $100,000 investment. Nupedia would sell ad space in its articles to make money. (So Bomis had 2 types of content sites to sell ad space - soft porn and scholarly articles).

Nupedia relied on highly knowledgeable authors (preferably PhDs) who were experts in their fields to write the articles. It coupled that with an extremely strict review process. There was a strict 7 step review process set up to ensure the accuracy of the content, with the articles reviewed by multiple scholars and professors in the process. This obviously took a lot of time. In fact, it took so much time that only 21 articles were approved and published on Nupedia in its first year. The extremely bureaucratic process made no business sense and resulted in Nupedia shutting down in 2003.

Luckily for Wales, the team at Nupedia had started another side project called Wikipedia in early 2001. The name came from the Hawaiian word "Wiki" which meant fast. And that was the idea behind the project - allow collaboration of different users on an article to push it out faster. Eligibility rules and review process were simpler than Nupedia's. Anyone could contribute instead of just PhDs. And their contributions would be monitored by other community members or contributors, hence ensuring accuracy doesn't take a hit.

The entire Nupedia community quickly transitioned to Wikipedia. And in just a matter of 4 years, Wikipedia had a repository of 2M articles.

Just like Nupedia, Wales initially wanted to make money from Wikipedia by placing ads on the website. Because that was what Bomis was about. Create content to attract users & show them ads (what news websites do as well). Even the web address for Wikipedia was www.wikipedia.com indicating commercial interests.

This however met resistance from their Spanish users who felt commercial advertising would harm the sanctity of the site's content. The Spanish contributors, in fact, left Wikipedia in 2002 to start an independent ad-free project 'Enciclopedia Libre' as a form of protest. Wales quickly reacted to this development by making Wikipedia also ad-free. In the same year, he also changed the domain from Wikipedia.com to Wikipedia.org.

But that isn't the end of the story. In 2003, a few of Wikipedia's servers (remote computers on which web pages are stored) went down creating the need for a replacement. This needed significant money which Wiki didn't have because...well, no ads. It was then that Wales ran the first donation campaign, targeting to collect $20K in a month from the community of users. Instead, he collected $30K in just 2 weeks. This infused confidence about the possibility of donations as a continuous stream of income hence making it a regular affair.

This raises an interesting question though. Initially, Wikipedia shifted to a no-ad model out of pressure and fear from the possibility of an alternate ad-free online encyclopedia being created. However, 2 decades into existence, Wikipedia is so big that it is almost impossible to create another one. So, what stops Wikipedia from doing ads now?

There are three reasons for this, Wales says.

Higher ad rates in developed countries
When a website wants to show display ads (those irritating ads on the side of news articles), it generally uses an ad network (like Google's) which provides it with ads. And the website gets paid basis two metrics: Cost per Click (CPC) and Revenue Per Mille (RPM). CPC is the amount the site is paid when a user clicks on an ad. RPM is the amount of money the site is paid when an ad gets 1000 views (called impressions). Taking the example of the largest ad network, Google Adsense, both CPC and RPM are much higher for developed countries than in developing countries. For example, the average CPC and RPM rates in the US are 4 times that of India.
Now consider for a moment that Wikipedia earns only via ads. It would want to receive more traffic from developed countries because of the higher CPC and RPM rates. Which means there would be more articles on places, events, personalities, companies, etc. from the developed world only. The bias would be so strong that they would prefer an article of insignificance from the developed world (eg: a page on Kim Kardashian's dog) vs. an article of significance from the developing world (eg: a page on Indian elections). This goes directly against Wikipedia's cause of 'sum of all human knowledge for everyone'.
Tending towards commercial topics only
The second form of earning through ads is affiliate links in articles. Websites can generate and embed a personalized Amazon link of a product in their related articles (eg: link to OnePlus phone in a phone review website). And if a visitor makes a purchase by clicking on that link, the website receives an affiliate commission from Amazon. Amazon is just an example. A lot of other e-commerce run affiliate programs too.
The obvious problem here is that few types of articles are much better fit for affiliates than others. For instance, an article on Air Jordan shoes is a better fit than an article on Schrodinger's equation (what do you sell on this page?).
If Wikipedia starts including affiliate links, they are incentivized write more on commercial products like iPhone, Tesla or a credit card vs writing on history, literature and maths. Just like the previous point, this will result in biased content.
Loss of community
This last one is pretty obvious. Wikipedia works because of its community of over 250,000 editors. Given that it is not-for-profit, the users consider it honourable to contribute to this entire mass of the knowledge for free. If Wikipedia were to switch to a commercial model, there is a high chance that the majority of its community will stop adding/editing information for free. Without this highly engaged community, a lot of content will become outdated after some point in time reducing the value of Wikipedia.

These points make a lot of sense (and are noble too) but put Wikipedia in a position of risk. It is solely dependent on donations as an income stream (kind of like us :P). But it has expenses to pay, like employee salaries and hosting expenses for its millions of web pages (think 'storage expenses'). So, does Wikipedia receive enough money to cover all its expenses and then some? Or is it really at the brink of death like the donation request at the beginning of this piece made us feel?

Let's find out!

As per the 2018-19 audited financial report of Wikimedia Foundation, Wikipedia earned $120M in donations from both individuals and corporates. Few of the corporate donors include Google, Microsoft, Apple, Salesforce, Bloomberg, Adobe, and Cisco - many of them regular donors.

Total expenses were close to $91M with nearly 50% of this going towards salaries of a 400 member strong team. This means Wikipedia netted a neat $28.6M in profit. Not very bad, sorry, extremely good for a company that depends only on donations. In fact, Wikipedia has managed to post a profit every year since 2006 thanks to the generous (and deserving) donations.

Being profitable for such a long time has allowed Wikipedia to amass a sizable asset base of $176M, held either as cash or some form of investment. With liabilities of only around $10M, Wikipedia has a cushion of 2 years even if donations stop completely (highly unlikely though).

On top of all this, the organization also launched a Wikimedia endowment fund in 2016 to ensure the continuity of its existence. The idea of an endowment fund is to collect a huge sum of money and invest it somewhere. And then use only the money you make from returns to meet all your expenses (the initial amount invested stays intact and continues generating returns). The goal of Wikimedia's endowment fund is to raise $100M in the next 10 years. The fund has already received multi-million dollar donations from Amazon, Facebook, Google, billionaire George Soros, etc.

Given the yearly profit, large asset base, and a growing endowment fund, I'd say Wikipedia is in a pretty comfortable position financially. That said, donations are still its bread & butter, and that drying up would mean pulling the rug out from under Wikipedia's feet.

I'd have normally ended the piece here but thought we can do an interesting thought experiment instead. So here we go...

How much would Wiki earn if they actually did ads?

We did a few back-of-the-envelope calculations, obviously assuming the user community continues to contribute articles to Wikipedia.

$567M from display ads in a year
How did we arrive at this? By multiplying the following numbers
- No. of page views Wikipedia gets in a year: 263B
- % of users without ad blockers: 80%
  Around 11% of the world's internet-connected population use adblockers. We took a higher number of 20% to be conservative.
- No. of ads per page: 3
  Like other prominent ad-based news/content websites, Wikipedia can place up to 3 display ads per page - on the top, left and right. In mobile devices, it can choose to display ads in the middle of the page as well to remain at 3 per page.
- % times an ad is actually shown in a spot that is available (called "Fill Rate"): 30%
- Revenue per 1000 ad views (RPM): $3 ($0.003 per ad view)
  RPM rates are higher for developed countries and for certain niches like finance, real estate, & tech vs. music & books. For simplicity, we took an average of $3. We also assumed an equal proportion of visitors from developed & developing countries.
Summary
263B x 80% x 3 x 30% x 0.003 = 567M
$131M from affiliate sales
As seen earlier, affiliate links in articles would be the second obvious form of monetization. We arrived at $131M by multiplying the following numbers, most of them being global averages.
- No. of page views Wikipedia gets in a year: 263B
- No. of articles with an affiliate link: 20%
  There was nothing to benchmark against. So we took a conservative estimate.
- % of users who will actually click a link after seeing it (called "Click Through Rate" or CTR): 2%
- % of users (from among who clicked the link) who will actually end up buying (called "Conversion Rate"): 5%
- Average value of the item purchased: $50
  Okay, this is a tough one. The value of a purchase can range from something as small as $1 for a chocolate to $1000 for an iPhone. For simplicity, we assumed a conservative average of $50.
- Average commission earned by Wikipedia on each item purchased: 5% of the purchase value
Summary
263B x 20% x 2% x 5% x $50 x 5% = 131M

Just these two sources put Wikipedia at potential annual revenue of ~$700M. It can top these by adding a premium membership (read without ads), a daily curated newsletter, and merchandise (cups, t-shirts, etc) which will take it closer to revenue of $1B. It's pretty incredible that Wikipedia/Wales has left this much money on the table to chase a nobler cause.

That's it for today folks. Ciao!

If you liked this Simplanation, consider buying us a coffee 🙂 (exactly like Wikipedia :P)
Due to payment gateway technicalities, we have to give different buttons for readers in India and readers in other countries.

Buy us a coffee (₹)

Buy us a coffee ($)

Simplanations

Discussion about this post