[Editor’s note: We updated this by adding comScore’s numbers, effectively pushing up the value.]
In recent months and weeks, we’ve examined:
- when Google could surpass Microsoft in market capitalization here;
- what Google could generate from YouTube by adding Ad Sense on the social networking video site here;
- what YouTube’s pre-Google acquisition revenues could be here;
- what MySpace’s search business was worth to a partner here.
Today, we look at what Wikipedia.org would be worth - assuming a major thing: if it weren’t a non-profit - as a free, ad supported going concern. We first apologize to the Wikipedia.org community who would view this as heresy or blasphemous. This is strictly for entertainment purposes.
We’ll base our valuation on:
- Wikipedia.org’s traffic
- current advertising rates and
- valuation metrics for publicly traded companies and recently acquired media companies.
Admissions in Shortcomings of Model
Allow us to first state that if Wikipedia.org were a for-profit entity, it would have never enjoyed the meteoric rise in user-generated content, enjoyed the free hyperlinks it gets from other sites, and as such all of the free traffic. But we decided to venture in unchartered territory and make a few simple assumptions and project a value for the free online encyclopedia. As well, we point out one thing: while offline editions of encyclopedias like Britannica did not carry ads and generated revenue from book sales, it’s worth noting that the venerable encyclopedia’s online edition carry ads. So while we are simply suggesting that it is not inconceivable for an encyclopedia like Wikipedia.org to carry ads online, we definitely acknowlegde that if it were to run ads and suddenly become a for-profit media property, there would be considerable revolt by its contributors, readers as well as partner sites (would there? Read on). And, we acknowledge that there must be something in its “charter” that forbids it from being a for-profit, but the question has been asked by some readers of this publication so we took it upon ourself to accept the challenge and come up with this evaluation. You will notice a lot of “we assume” in this analysis… what’s that saying? Don’t assume because you make an ass of u and me. We’ve heard of it, it’s funny… and now that I beat you to that, let’s get on with the analysis.
TRAFFIC
US-based Traffic
According to Nielsen Net Ratings, Wikipedia.org’s US-based Home and Panel traffic for September 2006 stood at 33.3M uniques and 630M pageviews.According to comScore Media Metrix, Wikipedia.org’s Worldwide-based Home and Panel traffic for September 2006 stood at 153M uniques and 2B pageviews.We won’t get into the merits and accuracy of NNR verus comScore Media Metrix or Alexa, we’ll accept them since investors and analysts regularly use these as “God’s word.” We also note that according to Alexa, 55% of its traffic goes to en.wikipedia.org (the English version) while the rest goes to international editions. Assume the 55% English language edition includes traffic from the UK, Australia, Canada, even India and other largely English speaking countries.
International-based Traffic
But for the sake of simplicity, assume 50% of Wikipedia.org’s traffic comes from the US, 50% from the world. So using these stats, using Nielsen NetRatings’ numbers, we project Wikipedia.org to have a worldwide audience of some 66M uniques and whopping 1.2 billion pageviews every month. To get an average between comScore and NNR’s numbers, we derive a worldwide unique count of 109M uniques and 1.6B pageviews.
Display Banner Advertising Impressions
To derive a number for ad impressions, simply multiply pageviews by two, one 728×90 leaderboard and one 250×250 billboard (or island) ad. This means that Wikipedia.org has over 3.2 billion advertising impressions per month to sell.
Wikipedia.org’s Banner Revenue
Allow me now to say that in my years as an online ad sales guy, I managed to sell out the site I represented but it was a much, much smaller site (5M uniques, 50M pageviews, 250M impressions) that Wikipedia.org’s inventory is a very large number of ad impressions for any salesperson or staff to sell, so we assume that Wikipedia.org only sells 75% of its ads (and it hits this mark by using ad networks but even they cannot in aggregate sell everything out).
So:
- 25% of its inventory sells for $2 CPM (Cost per Thousand of Impressions) and generate Wikipedia.org $1.6 million;
- 25% of its inventory sells for $1 CPM (Cost per Thousand of Impressions) and generate Wikipedia.org $800,000;
- 25% of its inventory sells for $0.50 CPM (Cost per Thousand of Impressions) and generate Wikipedia.org $400,000;
- and 25% of its inventory goes unsold. It could then be argued that if Wikipedia.org would open up its site to advertisers, it could generate $2.8 million per month off display/banners advertising.
Wikipedia.org’s Search Revenue
We’ve already looked at the difference in the propensity to search on a search property vis-a-vis a content website. On Google and other search engines, you can estimate that 100% of visits lead to a search query. Total clicks to both organic and paid results might be 95-100% of the total queries (assume up to 5% do not click on anything). On content sites, historically we have seen a 20% proportion between monthly uniques and total search queries. To clarify a point I made in the earlier post, I am not saying that 20% of unique users will search something on a property’s site search box, we are saying that generally speaking, in private and public discussions with people working in search or who work in publishing and have struck search deals, you can see a site approximating a 20% proportion between monthly search queries and unique users.
We would also project that as the unique users and pageview count increases, and in particular on a site like Wikipedia.org where there is excessive amounts of hyperlinks, the number might be lower. But, once again for the sake of this analysis, we project that 20% of the unique user number represents a fair estimation of search queries per month: so 20% times 109M, or up to 21M search queries per month. Note that this translates to 726,666 daily searches.
About 5% of clicks off a search engine go to paid results. This number can go from 1 to 30%, but for this analysis we will project search revenue using a 5% click through rate (CTR) to paid results.
According to Fathom Online’s Q1 2006 data, the average Cost per Click (CPC) were $1.39. Wikipedia.org can negotiate at least a 50% revenue share from a paid results partner, so it can generate: 21.8M search queries x 0.05% CTR x $1.39 CPC x 50% Revenue Share = $757,550 per month in search revenue. We’ll assume that since it is already heretic to run advertising on Wikipedia.org, there are no revenues from licensing of its content, or sponsorships, video pre-rolls or of course, pop ups and pop unders.
In a given month, Wikipedia.org stands to make $757K in search plus $2.8M in display banner ads, or $3,557,550. Projected over 12 months assuming no growth, that’s a whopping $42M per year in revenues. Yes, not too shabby. Of course, we stress some obvious realities: if Wikipedia.org served ads, only 1% of the sites that currently link to it for free would continue to link to it; this would pummel its PageRank strength (notice how often Wikipedia.org’s result is tops on Google). This would effectively reduce traffic and potential revenue potential.
COSTS
We’ll assume that like today, Wikipedia.org will not spend anything on marketing.
The bulk of Wikipedia.org’s expenses are related to tech (hardware, hosting and bandwidth). On the Wikipedia.org entry on Wikipedia.org, it is claimed that its 2005 4th Quarter expenses were $321,000, of which 60% or $192.6K were related to hardware, per quarter. Project that times four and it’s almost $800K per year. Note that Wikipedia.org’s traffic seems to have grown three-fold since 2005 Q4 (according to Alexa), but since a one unit increase in traffic does not lead to a one unit increase in bandwidth and IT costs, we cannot simply triple that expense. If we did though, it would be $2.4M per year now. For a second metric, let’s use YouTube as a gauge. According to comScore Media Metrix, in September, 2006, YouTube generated 66M worldwide uniques vs. Wikipedia’s 153M worldwide. Yet Alexa ranks YouTube as the 9th largest property and Wikipedia.org as the 15th. But since YouTube hosts video content and Wikipedia.org hosts text-based content, we’ll assume that Wikipedia.org’s bandwidth fees are at the very most 10% of YouTube’s: $100K per month, or $1.2M per year. Add in hardware and everything else and you’re looking at $2M per year in Tech costs. Average out the Wikipedia.org growth model IT cost estimate of $2.4M and YouTube-based model IT cost estimate of $2M and you get annual IT costs of $2.2M. Currently, Wikipedia.org does not pay its contributors.
That’s its appeal. But if it were a for-profit entity, I doubt many people would contribute. But, many would nonetheless: YouTube’s users bulked up its content for free and helped the founders cash in; Digg has seen explosive growth and it’s certainly a business. Point is: on this one, don’t make any assumptions. Here, it can go either way: you can assume that it can still pay $0 or start paying out to regular writers. Estimating how much that would be is hard, sort of. Let’s look at two metrics: Weblogs Inc. pays contributors $500 per month. Wikipedia.org has - according to its own entry - 27,000 users who edit at least 5 articles per month; there are 17,000 English users who edit on the English site. Digging further, we see that 4,000 people edit about 100 articles per month. Clearly, we can estimate that Wikipedia.org can “survive” with 1,000 actual writers who know what they are talking about. This is guesswork, but there could be 1,000 writers and 4,000 editors, over time there’s a churn where editors “graduate” to writers and some writers move on to find new hobbies. So say you were to pay the writers $250 per month, over a year, that’s 1,000 writers x 12 months x $250/month = $3M for content. Play with all the variables if you wish: double the rate, that’s $6M (though I doubt anyone contributes to Wikipedia.org for financial gain). Double the number of writers but leave the rate the same, you’re at $6M again. While many help out with Wikipedia.org pro bono, assuming it was a for-profit, it would probably have to pay for administrative employees (which we assume it does today anyway since only 60% of its budget is spent on hardware). Project $60K per year for 10 full-time people and you have $600K in administrative costs.
You tally all of the costs: + annual IT costs of $2.2M
+ annual administrative costs of $600K
+ annual content costs of $0, $3M or $6M and you are in the $3-10M ballpark for total costs.
The variable clearly is the content. But even if it were at the high end, you can see that with its enormous traffic and quality content, it could generate a yearly profit of $35 million! Of course, this is all for entertainment. So, we know the revenue and profit potential.
VALUATION MODELS
Publicly-Traded Company Basis
Looking at a basket of companies in the content, social networking, user generated content space that could be seen as reasonable comparables of Wikipedia.org, you get a P/E multiple of 30. Of course, you have to add a liquidity discount, so say 50% to be conservative, so a P/E of 15 times earnings. Multiple of 15 x a profit of $35 million = $560M valuation.
Recently Conducted M&A Basis
Last year, the average multiple acquiring companies paid was over 15.9 times EBITDA. This number goes up considerably, News Corp. for example paid 40 times EBITDA for IGN Entertainment, but IGN’s got a deep management bench, a long history of setting and meeting targets (I presume) and it has 300 employees. All to say, let’s be really conservative and estimate that given all of the factors and what if’s we outlined, Wikipedia.org would only get 15 times earnings, awfully close to the publicly-traded multiple after we applied a liquidity factor. That’s a multiple of 15.9 x a profit of $35 million = $600M valuation. Average out method 1 and 2 and you are looking at roughly $580 million market value for Wikipedia.org. Incidentally what News Corp. paid for MySpace’s Intermix last year. Again, we’re only having fun here, but while we’re at it, maybe just maybe we need to take a look at that Wikipedia.org charter again… If you would like other crazy valuation or off-the-wall estimates, email me at ash[at]mojosupreme.com
[Editor’s note: The most interesting thing about this analysis is that the first time we did this exercise, we used Nielsen Net Ratings number, which gave Wikipedia.org a lower unique count, as such, that gave us a valuation under $400M. Adding comScore Media Metrix, the value shot up to $600M, suggesting that indeed, valuations should probably stick to revenues and profits and not eyeballs…]
Part II - Why Wikipedia Should Avoid Ads (For Now At Least)
Subscribe:
October 27th, 2006 at 9:52 am
Interesting to see what Wikipedia could be earning. I will include this it in my newsletter. But it is not clear to me who you are. Referring to you as “some blogger” I prefer not to do. Is “froosh” Ashkan Karbasfrooshan? Greetings, Walter
October 27th, 2006 at 2:32 pm
[…] Will Google Become The Future of Online Audience Measurement? posted by froosh Three things: 1 - Must.Increase.Traffic Growing a website’s traffic is hard, much harder than it was in earlier years. I’ll write more about this later on. Like many people, I use Google Analytics (formerly Urchin) and look at it every day to see traffic trends on our media properties. We can see that traffic has been steadily growing day-after-day, month-over-month. Upon reading TechCrunch’s post on Digg being shortchanged by comScore Media Metrix and Fred Wilson’s defense of Media Metrix, I thought: “man, there must be a better way to do this.” There is. More on that in a bit. Mr. Wilson is right in thinking that comScore Media Metrix can’t be all that “flaky” since investors, media buyers and others use online audience measurement services like it (or Nielsen//NetRatings’). But, the fact is that these interested parties use these services because there is no better alternative and not because they’re all that great. In other words: if you are a large ad agency looking at buying ads on site A and site B on behalf of an advertiser, you need something to cover your derriere. Bottom line is that these services are great for relative analysis (site A vs. site B), but not absolute analysis (one site’s traffic). HitWise has also entered the fray, and while it’s a nice product I’ve used, I could not even tell you how it’s all that much better or different (note to self: research that). As online measurement services, both use a panel of surfers to project estimates of traffic. Even for large sites, the variance is massive. For example, according to NNR, Wikipedia.org generates some 60M uniques, according to Media Metrix, it generates 130M. I know what you’re thinking: “slight difference.” Imagine an investor that tries to estimate the value of a property using either service. They would miss by a mile. In our exercise where we tried to determine the value of Wikipedia.org, the value ranged quite a bit depending on which service’s traffic we used. This is obviously a problem. But, it gets worst. 2 - Alexa is one of the best and worst things online. Alexa is good because it provides a free and readily available gauge of a site’s traffic rank relative to others. But, Alexa is wildly erratic and erroneous: it’s biased towards techies who have downloaded the Alexa toolbar (which most anti-spyware programs view as a threat to your computer, by the way). All to say, Alexa is not quite representative of the broader Web universe, but neither are Nielsen//NetRatings, Hitwise or comScore Media Metrix. Allow me to vent a bit: our traffic has been growing slowly but surely since our launch in January 2006. We contact advertising agencies, potential investors contact us, and both stakeholders seem to look at Alexa. We give them our log stats (we’re still not getting the magical number to pop up on Nielsen//NetRatings or comScore Media Metrix - though we will soon) so Alexa is what many look to for guidance. Finally, we’re starting to gain some traction there, but even as the Alexa ranking grows, I make it a duty to explain to people why Alexa is not the solution. Google’s Foray in Analytics? I’ve always maintained that Google Analytics can make a dent against Media Metrix, Nielsen//NetRatings, Hitwise and Alexa because it’s not a panel, but rather, a site-specific measurement of traffic. Formerly known as Urchin, its competitors are numerous and include WebSideStory. Here is what Google should do: - Allow website owners to make public their Google Analytics traffic. Say we generate 1M unique users according to our third party internal logs but comScore Media Metrix and Nielsen//NetRatings give us 500K and 600K uniques respectively, as a publisher, I would gladly make public this number. Why wouldn’t I? Overnight, Google would become very relevant in the space and become more relevant than Alexa. - Google could then compile and aggregate the traffic rankings based on Google Analytics’s unique count and publish those, giving an extra incentive for websites to publish and adopt Google Analytics. Over time, more and more analysts, investors and ad agencies would look at Google Analytics and not the “flaky and not so flaky” sources. The best part is that once this adoption reaches most sites, the traffic we quote would be site-specific and not panel-based. This would make Google way too powerful for many to embrace, but they already own 50% of the search market, are the only company I can see having the resources to make this free and available to all. And, it’s not like the flaky and not so flaky online audience measurement services have too little power, now is it? There you have it. October 27th 2006 Posted to Uncategorized […]
October 28th, 2006 at 5:21 am
[…] Advertising on Wikipedia? HipMojo presents a thought experiment, using Wikipedia’s rising popularity according to the latest ComScore rankings to calculate the economic value of Wikipedia. […]
October 29th, 2006 at 12:47 am
[…] Note: I’m bringing this up again a good friend pointed me to this very, very conservative valuation of Wikipedia. Wikipedia if it was a private company would be worth $5B. […]
October 29th, 2006 at 1:26 am
Your Search click estimations are all wrong…I manage a big content website with a adsense search box which generates millions of queries each month, so i think i have better data to guestimate.
The CTR for search sponsored links you can get is close to 15% but the average CPC is less than 20 cents which means the effective eCPM wouldbe around $28. So a hypothetical wikipedia adsense search box would generate around $600k/month with 21 million queries per month.
I think if wikiepedia decides to go for ads it should opt for adsense text ads instead of a banner. The contextual ad system would work beautifully for wikipedia and can easily generate a CPM of more than $5.
October 29th, 2006 at 4:48 am
[…] Ok, I saw this post at Jason Calacnis’s weblog, the story ended up making digg. HipMojo projected that Wikipedia could make $42 million a year. No disagrement here, unlike Myspace, Facebook, and YouTube who have pretty mediocre traffic, Wikipedia has niches which can easily earn double to triple digit CPMs. […]
October 29th, 2006 at 11:39 am
[…] Note: I’m bringing this up again a good friend pointed me to this very, very conservative valuation of Wikipedia. Wikipedia if it was a private company would be worth $5B. […]
October 29th, 2006 at 1:09 pm
[…] I emailed Fred Wilson my analysis on Wikipedia’s potential value if it were a for-profit and if “many things could be assumed.” I figured that as a VC, he might have something interesting to say or simply enjoy the heretic exercise in putting a monetary value on something so cherished and beloved on the Web. […]
October 30th, 2006 at 3:03 am
[…] What is Wikipedia.org’s Valuation? Filed under: […]
October 30th, 2006 at 3:17 am
That was a great analysis.
October 30th, 2006 at 11:48 am
Larrying Wikipedia
If Wikipedia were a for-profit company, what would it be worth? It’s a question I’ve been thinking about - and, apparently, others have, too. HipMojo did some back-of-the-envelope calculations last Thursday and came with $580 million as an estimated …
October 31st, 2006 at 4:15 pm
[…] In our earlier analysis on Wikipedia.org’s potential valuation if it were a for-profit, ad-supported entity, we outlined that Wikipedia.org gets a lot of incoming links by virtue of being a non-profit. Web site operators are notoriously stingy with outbound links, that’s a sad state of affairs. When Google launched, Ad Sense did not exist (obviously) and as such, traffic in itself had little value. Today, with Ad Sense, traffic is automatically equal to revenue, so website operators tend to hesitate before linking to a website. As such, Google’s Pagerank tends to emphasize either a site like Wikipedia.org that is not at all commercial or pages on amazon.com whose affiliate program allows the linking website to get a cut of the potential transaction. This has been my number one concern for traffic flow on the Web since I began to work in the search engine industry in 2000. […]
November 2nd, 2006 at 1:21 pm
Thanks for your posting!
According to our technicans, we had about 2.5 billion page impressions (on wikipedia.org) in december 2005. This number was gathered using two different methods to estimate page impressions on our projects. Today, we have much more impressions per month for sure.
Tim (Wikimedia Support/Press Team)
November 5th, 2006 at 9:52 pm
[…] Frankly it can be seen as Wikipedia.org should Wikipedia.org become a for profit: good information on a topic, with ads (for an examination of what Wikipedia.org could be worth as a for-profit, click here; to see why Wikipedia.org should avoid ads altogether, click here). […]
November 7th, 2006 at 5:44 pm
[…] All to say, I knew I had to dig deeper. You will have to forgive me for not being very familiar with PopSugar.com, after all, I am certainly not in their target market. Given the pedigree of the VC in question, I had to dig deeper, but between analyzing the value of Wikipedia.org were it to be a for-profit, estimating when Google could overtake Microsoft in market value and managing the Mojo Supreme network of sites, investigating the “PopSugar rationale” kept falling on the backburner. […]
November 7th, 2006 at 7:24 pm
cool
November 10th, 2006 at 11:53 am
[…] Anyway, here goes: when I came across the war of words between Nick Carr and Mike Arrington, I initially did not know who was who. I found out about Nick Carr when he linked to a post I wrote on YouTube or Wikipedia some time ago. I had come across Michael Arrington’s TechCrunch a couple of months ago but had no clue what the site was about or how influential it was. In case you don’t either, Tech Crunch is one of the most influential sites out there and Michael Arrington is the man of the hour in the Web 2.0 blog space. Any single individual who can attract an audience in this cluttered space deserves credit. But that being said… […]
November 14th, 2006 at 11:46 am
[…] But, when you’re also a writer of a website that questions what Wikipedia should be worth as a for-profit, if Google can overtake MSFT in market cap, whether Google overpaid for News Corp.’s search business or if it’s the 21st century version of Standard Oil or Microsoft, you can’t help but ask the question: […]
November 16th, 2006 at 7:35 am
[…] Popularność Wikipedii tłumaczy dlaczego firma HipMojo zajęła się szacowaniem hipotetycznej wartości Wikipedii (gdyby była komercyjnym przedsięwzięciem utrzymującym się z reklam) - i wyceniła ją na 580 milionów dolarów. Jason Calacanis (CEO serwisu do blogowania Weblogs, Inc.) uważa, że encyklopedia jest warta 10 razy więcej - około 5 miliardów dolarów. Zdaniem Calacanisa sama firma AOL byłaby gotowa płacić 100 milionów dolarów za możliwość zaoferowania Wikipedii darmowego hostingu w zamian za malutką grafikę informującą o tym fakcie, umieszczoną na głównej stronie encyklopedii. Calacanis uważa, że grzechem jest tracenie każdego dnia “milionów dolarów, które mogłyby zmienić świat”. […]
November 17th, 2006 at 12:30 pm
[…] HipMojo.com - IT, Video, Web, Technology, Gadgets » What is Wikipedia.org’s Valuation? what Wikipedia.org would be worth… (tags: Wikipedia) […]
December 19th, 2006 at 1:01 am
[…] Wikipedia leaves $100M on the table (or “PLEASE Jimbo, reconsider–media philanthropy could change the world!”) Update: Story is now on digg. While on the subject of media philanthropy….I sat next to Jimbo at a Wikipedia dinner over the summer. I begged him to put a leaderboard on Wikipedia and told him I would get AOL to sell it and host Wikipedia–for free. He declined saying there will never be ads on Wikipedia. I then explained to him in detail how that one leaderboard could make over $100M per year. I told him that they should take the $100M and give it to charity. They could help fund MediaWiki, the EFF, Firefox, and dozens of other open source projects. Jimbo: please reconsider!!! I know I can get AOL to sell the inventory at zero cost to you guys and we will donate the bandwidth. Just give us a little 25×25 pixel thank you (i.e. “Hosted donated by AOL.” That’s it.Note: I’m bringing this up again a good friend pointed me to this very, very conservative valuation of Wikipedia. Wikipedia if it was a private company would be worth $5B. […]
April 17th, 2007 at 12:06 am
How worthless is this “valuation”? Or is this just another PR move since it is being quoted by other sites? As originally pointed out, it is ONLY where it is because of hundreds of thousands of UNPAID volunteers and no ads. If Wikipedia were to lose this its valuation would drop to nothing. Another dot com/org bust.
April 20th, 2007 at 3:45 am
[…] Popularność Wikipedii tłumaczy dlaczego firma HipMojo zajęła się szacowaniem hipotetycznej wartości Wikipedii (gdyby była komercyjnym przedsięwzięciem utrzymującym się z reklam) - i wyceniła ją na 580 milionów dolarów. Jason Calacanis (CEO serwisu do blogowania Weblogs, Inc.) uważa, że encyklopedia jest warta 10 razy więcej - około 5 miliardów dolarów. Zdaniem Calacanisa sama firma AOL byłaby gotowa płacić 100 milionów dolarów za możliwość zaoferowania Wikipedii darmowego hostingu w zamian za malutką grafikę informującą o tym fakcie, umieszczoną na głównej stronie encyklopedii. Calacanis uważa, że grzechem jest tracenie każdego dnia “milionów dolarów, które mogłyby zmienić świat”. […]
June 6th, 2007 at 8:48 am
[…] Note that if Wikipedia was a private company, it would be worth billions. […]