I included this chart in my post yesterday.
I selected it because it made the point I wanted to make (that delicious was languishing under Yahoo!’s ownership). I did check the chart against several other tracking services and was confident it was correct. And it is, sort of.
I got the following email from Joshua yesterday afternoon:
We continue to grow normally.
Unique users is not a good measure of our growth, though.
Much of our traffic is through the firefox and other browser extensions, which is not measured by these systems.
Additionally, we cut off search indexing several months ago, which also hurts the UU numbers.
So I got it wrong, at least partially. And I want to correct the record.
Delicious continues to grow under Yahoo!’s ownership.
I find Joshua’s first point particularly interesting. Much of delicious’ traffic is through firefox and other browser extensions. That rings true to me because it’s how I use the service.
I wonder how many other web apps are accessed via third party services (twitter’s traffic is largely through its api)?
And if that’s a growing trend, then what does that mean for our ability to measure audiences, traffic, and growth from a distance?
Its a very valid point – I use delicious through firefox too. Its interesting that delicious seems to have escaped the worst of yahoo’s ID system. I tried signing up a friend to mybloglog, the sign up is now a total nightmare (for a non yahoo user) since the takeover same with flickr.
not being able to measure audiences is yet another reason why advertising as a revenue source for online enterprises will become increasingly iffy … alternatively, that would be a good niche for a new business
Outside of pure website/page analytics, it means we don’t have much ability to measure audiences, traffic and growth from a distance for many things. delicious, streaming video, things done via SMS — I’m sure the list gets fairly long of stuff lots of people do that can’t easily be tracked from the outside looking in.
This has been a growing problem. Neilson and Comscore have been working to address this, without much real success. The services themselves can track this, but it is not independent or particularly verifiable.Some of the enterprise web analytics providers offer back-end analysis packages that could probably be leveraged to report externally, letting a third-party verify things. There may be a business here…
These kinds of tracking numbers are meaningless in our business too.You only need to come to our website once. Many of our subs sign up over the air. Many of our subs sign up when a friend forwards them a show (they NEVER come to our site).We track listener sessions a day. SMS messages sent per day. We track lag time between SMS notification and when a sub listens to the show. We track detailed stats about listener behavior during a call. Those real metrics have very little correlation to traditional “web traffic”. These real metrics have much more detail and value than web traffic.
uniques may be growing normally through those other channels but I wonder what’s happening to engagement/usage. I definitely bookmark a lot less to delicious than I used to (many other channels for saving/sharing content – tumblr, facebook, twitter, …). I suspect you do as well as do many other long-time users of the service.would be interesting to see a graph of bookmarks/week over time
Wow. I, and everyone I know, use the Firefox extensions. Is the failure to measure a technical hurdle or just a procedural issue?
Its a problem not limited to just accessing via third party services. Analytics just does a bunch of manipulation on web server logs, and web server logs are based on URL activity. As web sites becomes “applications” with ajax and other technologies, what goes on inside of those apps is not URL activity. Its data activity and non-URL user events. So it becomes a problem of garbage-in, garbage-out.
I am not sure that ‘Twitter’s traffic is largely through its API’. ReadWriteWeb recently reported (http://www.readwriteweb.com… that, despite all the 3rd party apps, 56% of twitter traffic is still via the web.
that’s posting. but most of the views come through the api
No one pointed this out, but interesting that just a few days ago Seth Godin talked about FireFox and how people that use FireFox are a “market segment” unto themselves. He equated the browser to a gateway drug.Full article here:http://sethgodin.typepad.co…
When I did my analysis of Friend Feed services, bookmarking was an even split between Google Reader shares and Delicious.I’m seeing less Delicious use then their used to be, but more Google Reader shares.
Why would del.icio.us cut off search indexing? I would think Joshua would be delighted if the “funny-video” tag was the first search result at Google for the search team “funny video”… it would mean both greater distribution and influence.
Unless Yahoo views that it is advantageous to stop non-Yahoo search engines from indexing Yahoo content like delicious. The theory would be that if Yahoo’s search can find stuff buy Google can’t, it’s advantage Yahoo.Kind of a scary game if that’s what’s going on, no?Best,BW
I was also baffled when I saw that “cut off search indexing” line.But I’m not sure they are blocking Googlebot. Check out “http://del.icio.us/robots.txt”. I see this:User-agent: GooglebotAllow: /Also, try this query: “site:del.icio.us’. There’s a bunch of pages indexed. Even though most of them seem to be profile pages (and not tag landing pages), Google’s Site query is notoriously and intentionally incomplete. If you don’t block Googlebot via robots.txt, you can also block them with “no follow” tags. A quick look at Delcious’ home page source doesn’t show no follow tags pointing to the tag landing pages. For example, I see this:<li><h4>fun</h4>Looks indexable to me.And they’re certainly not blocking the Yahoo spiders. Check out this query on Site Explorer – 58M pages indexed.:https://siteexplorer.search…I’m not a full time SEO (I just play one on TV) so I might be missing something…. but Del.icio.us looks fully indexable by each of the major search engines to me.
@donaldfoss – I can certainly attest that there is a business here…it’s called Mashery. We have dozens of customers who use us to manage their APIs, and one of the most important services we offer is reporting on API usage and activity. @DanMalven has it absolutely correct – “web analytics” provide either server log analysis or browser activity logs, but if you are not accessing an API directly from the browser (and even if you do access from the browser, but do so through a widget), there is an awful lot of activity data that are not captured by web analytics packages. That’s why we built this company, and why our customers pay us each month.Oren MichelsCEOMashery
I almost never access delicious via the home page. It’s primarily via Firefox or via Lijit
How does this impact revenue? Am I over simplifyin’? If folks don’t actually visit the site, can get just the content delivered to them and can use the service, features with out actually engaging the site, how is money made, in a ad revenue model? I MUST be missing somethin’. I don’t like is when I’m missin’ somethin’!
Doesn’t this fit in with microsoft’s strategy of connected apps (which i assume still means the ability to access via a browser) as opposed to everything in the cloud (Google’s strategy)?