The biggest surprise fro me in the social media rankings that I posted earlier today was the appearance of document sharing service Scribd in the top 20. According to the comScore numbers, it has more unique visitors worldwide than imeem and almost as many as Bebo, with 23.5 million visitors in November, 2008. (In the U.S., it had about 4 million visitors). This is a serious accomplishment for a startup that launched less than two years ago with $300,000 in seed capital.
Scribd, which raised a $9 million B round earlier this month, grew 218 percent from November, 2007. Scribd is heads and shoulders above other document-sharing services such as Docstoc (1.6 million uniques) and Issuu (1 million) in terms of attracting users. (All three are excellent services, by the way, and Docstoc is much closer to Scribd in the U.S.). Scribd users upload 50,000 documents every day.
What this tells me more than anything is that the concept of document sharing on the Web has legs and there is a real demand for it. (Unless Scribd’s numbers are artificially inflated—Quantcast shows an unusually steep drop-off in December, but that could just be a problem with Quantcasts’s data). Desktop-bound document formats like PDFs, Word docs, Powerpoint slides, and spreadsheets are increasingly irrelevant if they cannot be viewed and shared online directly in a browser. Scribd’s paper is an embeddable Flash viewer for nearly all document types. (I’ve embedded a document with sample job interview questions below).
It looks like we are not the only ones who like to embed documents on the Web.
Update: CEO Trip Adler explains the Quantcast discrepancy:
The drop-off on Quantcast was caused by the fact that we removed the Quantcast pixel from Scribd for a little while because it was slowing down page loads. However, I’m not sure why these numbers are still on the low side. According to our Google Analytics, we did 39.9 million unique in November, and if you include views of iPaper on other websites, it was way over 50 million uniques.
E pur se move. By that I mean, it still doesn’t and cannot monetize. So who cares? Another Y Bomb.
Galileo, here is how DocStoc and Scribd will monetize:
http://tinycomb...-for-documents/
They should be doing pretty good with adsense alone since it has millions of text based pages on specific niche topics it will match relevant adverts well.
Adsense is not enough for them either company to cover costs.
It’s harder to get that kind of traction than to monetize it.
Everyone I show it to loves Issuu. I put my resume/portfolio up there, and it’s a great way to share mockups with clients etc. Their sharing/embedding options are pretty well thought out, too. I’d say Scribd is better for plain old text, like what Erick posted, but Issuu looks pretty snazzy for visually richer docs.
Your credibility is undermined by your use of “rich” in reference to media.
…and we are supposed to give credibility to you as a grammar troll?
I’ve been following scribd ever since they launched and have uploaded/downloaded several documents. It’s an excellent service that always had the potential to be big, the team is dedicated and focused on enhancing the service, that makes a recipe for success.
Personally I like docstoc because of their ability to keep documents private and yet embed them. Makes for better control on your own website.
So, what % of Scribd’s traffic was due to TC referrers? You have both sets of data.
Watching for the web document to become so much more in 2009 — we can view documents online and embed them, but what will be interesting will be the set of services around those documents.
There are some interesting plays to be made around:
1. enterprise systems such as visual document management
2. the truly virtual document, embedding several different file formats together
3. the data mined from all those uploaded documents
Didn’t Web 2.0 end? What was the business model here? And can you really call it a company on 25K unique visitors a month? And much does it cost to serve and store all that stuff?
Are we talking sock puppet business model here?
We already know the haziness under which TC promotes companies, so would be good to know how much was the TC manipulation.
You lost me at 25K.
Uhh… I think a more careful reading of the graph is in order. That was 25K thousands. 25 million in common parlance.
some facts:
-90%+ of Scribd’s traffic is transient search engine traffic (it’s a pure SEO play)
-Their bounce rates are an atrocious 80%+
-There is no engagement in building a social document sharing site (by and large, it’s a repository of pirated/copyrighted docs) in contrast with say a social video sharing community like Youtube
-Scribd’s meteoric growth gives them a *VERY* short window of opportunity to use this veneer/hype of being marketed as a legitimate service to cash out before they get litigated into oblivion or have to settle for the voluminous amount of pilfered material on the site (much like Google was forced to in October 2008).
-Scribd doesnt actually remove all copyrighted docs or adult material from the service for the purposes of SEO. They merely ‘hide’ it from being searchable on their site. (So 85-95% of their traffic is being sent to either copyrighted or adult related material like erotic sex stories)
-The latest moves to bring in the Bebo exec (and heck, this post itself) is just last ditch windowdressing before the eventual pump and dump.
caveat emptor
Another guy writing about what he doesn’t know just because he feels like being negative. I used Scribd and uploaded some copyrighted material which I thought would not be detected but the document was automatically removed. I was alerted by email as follows:
We have removed your document “_____” because our text matching system determined that it was very similar to a work that has been marked as copyrighted and not permitted on Scribd.
Like all automated matching systems, our system is not perfect and occasionally makes mistakes. If you believe that your document is not infringing, please contact us at copyright@scribd.com and we will investigate the matter.
As stated in our terms of use, repeated incidents of copyright infringement will result in the deletion of your Scribd.com account and prohibit you from uploading material to Scribd.com in the future. To prevent us from having to take these steps, please delete from scribd.com any material you have uploaded to which you do not own the necessary rights and refrain from uploading any material you are not entitled to upload. For more information about Scribd.com’s copyright policy, please read the Terms of Use located at http://www.scribd.com/terms
I doubt that traffic is the most relevant way to measure the success of this kind of website.
I would much rather value average reader per document, number of documents uploaded per month, time spent per document, number of documents downloaded, embedded, etc. All this trended and segmented (e.g. without porn).
At least these KPIs measure engagement of some sort and bots do not engage!
Too bad all of their content is “stolen”.
lol, one mans “stolen” is another mans “borrowed” =)
hey erick,
i suppose the quantcast decline is because scribd stopped tracking. perhaps the daily chart is a more clear picture:
http://www.quan...ibd.com#traffic
(then click on people per day)
happy holidays.
dk
Mispelled word in first sentence.
These posts about sites doing well sure bring out the ass clowns!
ComScore = Garbage sample based data.
Quantcast Quantified = A public version of Google Analytics
Quantcast Quantified should be the only source you guys trust because it uses a javascript pixel load like Google Analytics. This means it is real numbers minus the 2-15% of folks that might have javascript turned off (depending on the site). So, call it accurate -5%.
I’ve been in this business since before ComScore existed and the fact that anyone uses their garbage numbers based on some b.s. sample is crazy.
TechCrunch should really ban using ComScore’s numbers–which most folks think are bought and paid for–and move to quantified numbers.
If you won’t quantify your site with Quantcast than your numbers are fake. Alexa and Compete are gamed to hell.
The only thing that matters is google analytics and Quantified Quantcast sites.
end rant.
What percentage of websites have been “Quantified” by Quantcast? It wasn’t many, last I looked. If it truly comes down to Quantcast to broadcast, what is your recommendation for performing comparison traffic analysis when one or more sites in your sample are not Quantified?
with TechCrunch’s influence they could easily push folks to quantify.
would be better for the industry as a whole to stop playing the fake, pay for play games.
Quantcast is not as consistent as comScore yet because not all sites are Quantified, including some of the most important ones like Google. Every measuring service has its problems, but at least comScore is consistent and comparable across a broad range of sites. (That said, it’s better for bigger sites than smaller ones).
What about Google Trends for websites. Do you trust those numbers?
http://trends.g...=all&sort=0
Isn’t this more about the homogenization of content? Bring all types of content into a single viewable format for the user regardless of how it initiated; a common denominator if you will? Most content is accessed as-is and for reading/viewing purposes, so it saves the creator the trouble of distributing it in a more common format for the world.
Another example might be the translation of all literature into English.
I like this blog.
Yeah, very useful blog.
I have found copyrighted e-books on Scribd that the authors are trying to sell for cash on their own site.
It’s a similar situation (though currently smaller scale) to the music business and this punches a hole in the advice (advocated by Arrington) that bands let music go free and make money on concerts and t-shirts.
How should an ebook author make money once their work is available for free on Scribd? T-shirts? Speaking tours?
With any serious marketer, ebooks are only a front end product where they give away a huge percentage of each sale to build a list of targetted buyers in a particular niche.
There are ebook publishers that pay 200% to affiliates – twice as much as the customers pay for the ebook.
They make the money back on backend sales.
The loss for such a publisher isn’t necessarily the money from the ebook sale, but the backend. Whether people who “find” a pirate copy of an ebook will become buyers of backend products is dependent on the niche, quality and type of the backend.
Just as with music where the real money might be in merchendise or concerts, the real money in ebooks is an upsell to a subscription site, service, physical product such as a home study course, consulting, or seminars.
Ummm…check out the “also searches for” and “also visits” sections under the lifestyle tab in quantcast http://www.quan...d.com#lifestyle
Not what I would call high quality traffic.
That is fantastic growth… hopefully they don’t run out of cash before they monetize the site and turn a net profit.
Great job on the growth regardless!
Jon
http://WoodMarvels.com – Create Unique Memories
Scribd is doing well, am sure.To hell with the different graphs which mean diff things.
The ipaper format is real good and user friendly.
Think they should improve on iPaper and make it possible to preserve animations,audio and effects in the uploaded documents esp. powerpoints.
I’m using the service for different types of documents and it’s great and easy to use. i really think that when they start adding ads – they might lose me…
It the documents like PDF files and PowerPoint presentation can not be viewed in its original format the service will be of no use. We will be able store only but to view we have to download them on desktop and that too if we have the relevant software installed.
Isn’t this more about the homogenization of content? Bring all types of content into a single viewable format for the user regardless of how it initiated; a common denominator if you will? Most content is accessed as-is and for reading/viewing purposes, so it saves the creator the trouble of distributing it in a more common format for the world
would be better for the industry as a whole to stop playing the fake, pay for play games.
The loss for such a publisher isn’t necessarily the money from the ebook sale, but the backend. Whether people who “find” a pirate copy of an ebook will become buyers of backend products is dependent on the niche, quality and type of the backend.
The loss for such a publisher isn’t necessarily the money from the ebook sale, but the backend. Whether people who “find” a pirate copy of an ebook will become buyers of backend products is dependent on the niche, quality and type of the backend.