Many of our keen members observed that late last week, Linkscape's index updated (this is actually our 27th index update since starting the project in 2008). This means new link data in Open Site Explorer and Linkscape Classic, as well as new metric data via the mozbar and in our API.
Index 27 Statistics
For those who are interested, you can follow the Linkscape index update calendar on our API Wiki (as you can see, this update was about a week early).
Although we've now crawled many hundreds of billions of pages since launch, we only serve our uber-freshest index. Historical data is something we want to do soon - more on that later. This latest index's stats feature:
- Pages - 40,152,060,523
- Subdomains - 284,336,725
- Root Domains - 91,539,345
- Links - 420,049,105,986
- % of Nofollowed Links - 2.02%
- % of Nofollows on Internal Links - 58.7%
- % of Nofollows on External Links - 41.3%
- % of Pages w/ Rel Canonical - 4.3%
These numbers continue the trend we've been seeing for some time where internal nofollow usage is declining slightly while rel canonical is down a bit in this index but up substantially over the start of the year (this likely has more to do with our crawl selection than with sites actually removing canonical URL tags.
Comparing Metrics from Index to Index
One of the biggest requests we get is the ability to track historical information about your metrics from Linkscape. We know this is really important to everyone and we want to make this happen soon, but have some technical and practical challenges to overcome. The biggest of which is that what we crawl changes substantively with each index, both due to our improvements in what to crawl (and what to ignore) and with the web's massive changes each month (60%+ of pages we fetched 6 months ago are no longer in existence!).
For now, the best advice I can give is to measure yourself against competitors and colleagues rather than against your metrics last month or last year. If you're improving against the competition, chances are good that your overall footprint is increasing at a higher rate than theirs. You might even "lose" links in a raw count from the index, but actually have improved simply because a few hundred spam/scraper websites weren't crawled this time around, or we've done better canonicalization with URLs than last round or your link rotated out of the top of a popular RSS feed many sites were reproducing.
Measuring against other sites in your niche is a great way to compare from index to index
If you've got more questions about comparisons and index modifications over time, feel free to ask in the comments and we'll try to dive in. For those who are interested, our current thinking around providing historical tracking is to give multiple number sets like - # of links from mR 3+ pages, # of links from mR 1-3 pages, etc. to help show how many "important" links you're gaining/losing - these fluctuate much less from index to index and may be better benchmarking tools.
Integration with Conductor's Searchlight Software
SEOmoz is proud to be powering Conductor's new Searchlight software. I got to take a demo of their toolset 2 weeks ago (anyone can request one here) and was very impressed. See for yourself with a few exclusive screenshots I've wrangled up:
And at the bottom of the series is Seth Besmertnik, Conductor's CEO, during the launch event (note the unbuttoned top button of his shirt with the tie; this indicates Seth is a professional, but he's still a startup guy at heart). Searchlight already has some impressive customers including Monster.com, Care.com, Siemens, Travelocity, Progressive and more. I think many in the SEO field will agree that moving further into software is a smart move for the Conductor team, and the toolset certainly looks promising.
Conductor's also releasing some cool free research data on seasonality (request form here). Couldn't resist sharing a screenshot below of the sample Excel workbook they developed:
mmm... prepopulated
SEOmoz's Linkscape index currently powers the link data section of Searchlight via our API and we're looking forward to helping many other providers of search software in the future. We're also integrated with Hubspot's Grader.com and EightFoldLogic's (formerly Enquisite) Linker, so if you're seeking to build an app and need link data, you can sign up for free API access and get in touch if/when you need more data.
The Link Juice App for iPhone
We're also very excited about the popular and growing iPhone app - LinkJuice. They've just recently updated the software with a few recommendations straight from Danny Dover and me!
The LinkJuice folks have promised an Android version is on its way soon, and since that's my phone of choice, I can't wait!
If you've got an app, software piece or website that's powered by Linkscape, please do drop us a line so we can include it. I've been excited to see folks using it for research - like Sean's recent YOUmoz post on PageRank correlations - as well as in many less public research works.
Oh, and if you somehow missed the announcement, go check out the new Beginner's Guide to SEO! It's totally free and Danny's done a great job with it.
Awesome post. Conductor is super excited to partner with the Moz team and by the incredible power of the Linkscape data. Our enterprise customers really love the integration and we're looking forward to watching the APIs evolve in the coming years.
As for the tie comment, it's partially the startup guy in me and partially the neck expansion I've realized as an outcome of being a "startup guy"...I was going to drop the suit...but since we had 125+ brands come to the launch event, I figured we needed to "look professional".
Thanks for the update. I am interested in the historical tracking and you have made some valid points.
I cant wait to see what you come up with for tracking.
Is there any time frame whatsoever as to when something may appear (maybe in the labs)?
And I am glad you pointed out the tie with the open button... I gain knowledge from every one of your posts.
:)
Unfortunately, it was a project Nick was working on before his departure. Current estimate would be Q3 of this year for historical data in Labs, and Q4 before it gets rolled into the full product/platform.
Luckily, we will be going back in time, so you should get to see link data from 6-12 months ago once it launches (this required some API processing/storage work, but my understanding is that's relatively on track).
This better happen!
Although not a current Searchlight customer, we've had a relationship with Conductor for nearly two years and have benefited from it tremendously. Way to go, Seth.
@Rand your team is really doing a good job at making this data more accessible outside SEOmoz so can we please flag the need for more international sites again :)
Yes definitely. You'll note that this index contains the largest domain diversity we've ever had - 91 million unique root domains. We are certainly working towards a more globally comprehensive set of sites.
If you have specific sites/pages you were hoping to see data for that we don't have, please drop us a line so we can improve!
Awesome!
Is SEOmoz keeping all the old data some place and not just replacing it? That way when you get it worked out, you still have all the historical data since you introduced the service.
Yeah - that's exactly right. We store the data on old indices now but haven't made them available yet. Our plans are to eventually provide a system that serves historical data along with current stuff, hopefully as early as Q3/4 of this year.
I'm excited to see these announcements about an ever expanding ecosystem around Linkscape. Nice work mozzers and partners!
Nice new tool.
Now I get to go re-run all of our reports. Maybe I can make it so they are all "compiling" while the US-Algeria World Cup match is on.
thanks for the update Rand and the good work keeping the index valid. Did not know about searchlight - I will definitely take a look.
Thanks for the shout out Rand. Looking forward to contributing many more in the future!
When creating the Link Juice App, we had the pleasure of working with Nick Gerner while he was still at SEOmoz.
The API really is very good. Thanks to Rand and the team for a fantastic product.
We hope fellow mozzers like the Link Juice App - we're looking for feedback as we're actively building the next version - let us know here or via our contact form.
Thanks
Charlie (Link Juice App)
P.S. If you'd like to know more about the android version, you can sign up to updates via our homepage
Love the app!
Thanks, Rand & team. I was honestly using the SEOMoz tools pretty sparingly until OSE.org came along, but since then, I am using it as a required part of my day to day. Well done.
Thanks for the update but will the link juice app available soon for Android ? :)
Sorana,
We've put up a form on the home page:
https://www.linkjuiceapp.com so you can't get notified when the Android app is ready.
We're also looking for beta testers.
Thanks
Charlie