Open Site Explorer.. Bit of a let down?
-
Hi, not too sure if this is a discussion or rant!?
I’ve been following SEOMoz for a couple of years now. Testing their tools, reading their blogs, sharing their content, and watching the whiteboard Friday videos (massive thumbs up to that one!). They are at the top of the game, no argument there.
Although I should have done so much earlier, I have eventually signed up to the pro version and am about to migrate all my clients over. But there is one major caveat which I’m not sure should exist.. Open Site Explorer.
Don’t get me wrong, OSE provides invaluable metrics, some that no other ‘crawler’ provides, but it is far from perfect, there are things that we SEO’s need, and are hoping to get!
Some feature SEOMoz could add;
- Ability to view a chart of ‘link types’ (i.e. blog post, social media, Press Release etc..) – Linkdex do this!
- Utilise a ‘fresh‘ backlink index as Majestic SEO do. ( we do use majestic alongside SEOMoz)
- Crawl more frequently – enough said!
- Index all ‘not so good’ backlinks – this will help identify what backlinks AREN’T helping.
I realise this is a lot easier said than done, and I’m sure SEOMoz are working on solutions.. just can’t wait till they launch!
What do you think? IS OSE more than it’s cracked up to be? Could it be improved? Let me know
Lee
-
Wow Rand, thanks for such a detailed and frank answer, I now have a better understanding of the issues you face.. it's much more complex than I imagined. Great to see that you''re doing all you can, look forward to the improvements.
Have a great week! Lee
-
Hi Lee - thanks for bringing these up! I'll try to answer with regards to each of the items you've mentioned:
#1 - This is very tough at our scale (literally 1 trillion+ links, and in the latest index about to launch, 150 billion+ pages across nearly 200 million domains). However, you're correct that we're working towards a classification system that will lean on some seeding + machine learning + user input. I'd suspect 12-18 months before we can launch it, though.
#2 - This is not currently in the plans, at least not at full web scale. We will have Blogscape back up and running soon, and that crawls ~10mm+ fresh sources daily, so if you have a link/mention on a blog/forum/news site with any notoriety, we should be catching those and updating hopefully every 6-8 hours. Mozscape (aka Linkscape, our full web index), will continue to be at least 2-4 weeks between updates and will require 3-4 weeks of processing. We're trying some new things on the technology side, but it's a huge challenge to get to Google's scale and keep our metrics, sorting, views, etc. Majestic can do this with their index because they simply export links directly into the consummable portion of the app, but it limits the sorting/views/filters and ability to generate high quality metrics like PA/DA/mozRank/etc.
#3 - Yup. Definitely agree and working on it. We're trying some new hardware stuff, new parallelization of processing and everything under the sun to make it work.
#4 - Our next index will probably have a lot more of this (and we'll be watching carefully to see how this performs for our customers). We're also working to build a spam score that maps against what we've seen Google penalize/ban over time (currently doing a bunch of research there), so you can get a good sense of what things are likely to be hit over time.
Mozscape/OSE is a huge priority for us and we have 7 extremely talented folks working night and day (and a lot of weekends) to improve this. In the next 3-6 months, it will get massively better than it is today - the next new index is only a few days away now, and after that, we shouldn't ever have such a long period without an update (this round was an impossibly hard-to-fathom confluence of problems that we've taken many steps to prevent from ever happening again).
Thanks for your patience and the good questions.
-
Ignore the last reply John, have just tested Link Detective on a site that I know the link backlink profile of.. it's massively inaccurate! The tool simply isn't able to identify correct link types as they suggest. Dissapointing!
-
I've read they would be indexing 2 x the data Anthony, not 3 x. Suppose there must be some confusion out there, fingers crossed it's 3 x the data!
-
Cheers John, will take a look when I get the chance.. have you used it before? Is it accurate?
Enjoy the weekend! Lee
-
Thanks for all the feedback guys, would be good to hear SEOMoz's official stance on this!
-
You might find this tool helpful to analyse your downloaded report from OSE http://www.linkdetective.com/
-
I have to agree with Kevin on this one...
Items 2,3,and 4 are days away.
Does it suck that OSE's Linkscape Index hasn't been updated since the end of February? YEP!
Will it be worth the wait next week when it is updated and contains three times the data as the previous crawls? YEP!
I consider SEOMoz a community, and though Linkscape's stagnant data is causing some issues, I realize that once those issues are resolved that I'll be getting more bang for my buck than I did before, and I'm totally willing to struggle for a month in order to have better tools at my disposal for many months to come.
Rand and SEOMoz are drastically improving the volume of data that we will have at our disposal. That takes time and resources. They aren't raising prices or making excuses, they're working out bugs and improving the service. I'll take it.
Anthony
-
Lee,
First, I need to say I OSE - it's invaluable in my forensic audit work. Having said that, I would LOVE to see the "link types" chart. The biggest challenge I expect they'd face in that is properly detecting where a link is on a page, or what class to put it in due to how unspectacular code consistency really is across the vast web. For example, I've seen others attempt this and fail miserably ( what should be identified as a sidebar (blogroll type) link is often seen as something else, and what should be seen as a main content area link is often seen as a sidebar link (too many WP themes for example, label their main content area divs as "sidebar2" for example).
And that just gets worse / more challenging when the need comes in identifying if a site is really a blog, or some other site that happens to use WordPress as it's platform.
Having said all that, I would be in heaven if the Moz team could get that data into OSE
-
From following all the talk, I'm pretty sure they are working on points 2, 3 & 4.
While all that is good, I personally feel they should concentrate on updating more often first, then move towards giving those extra features. The two month stretch between updates has been a real pain in the ass. But I'm sure they know that.
Browse Questions
Explore more categories
-
Moz Tools
Chat with the community about the Moz tools.
-
SEO Tactics
Discuss the SEO process with fellow marketers
-
Community
Discuss industry events, jobs, and news!
-
Digital Marketing
Chat about tactics outside of SEO
-
Research & Trends
Dive into research and trends in the search industry.
-
Support
Connect on product support and feature requests.
Related Questions
-
Our crawler was not able to access the robots.txt file on your site.
Good morning, Yesterday, Moz gave me an error that is wasn't able to find our robots.txt file. However, this is a new occurrence, we've used Moz and its crawling ability many times prior; not sure why the error is happening now. I validated that the redirects and our robots page are operational and nothing is disallowing Roger in our robots.txt. Any advice or guidance would be much appreciated. https://www.agrisupply.com/robots.txt Thank you for your time. -Danny
Moz Pro | | Danny_Gallagher0 -
Large site with content silo's - best practice for deep indexing silo content
Thanks in advance for any advice/links/discussion. This honestly might be a scenario where we need to do some A/B testing. We have a massive (5 Million) content silo that is the basis for our long tail search strategy. Organic search traffic hits our individual "product" pages and we've divided our silo with a parent category & then secondarily with a field (so we can cross link to other content silo's using the same parent/field categorizations). We don't anticipate, nor expect to have top level category pages receive organic traffic - most people are searching for the individual/specific product (long tail). We're not trying to rank or get traffic for searches of all products in "category X" and others are competing and spending a lot in that area (head). The intent/purpose of the site structure/taxonomy is to more easily enable bots/crawlers to get deeper into our content silos. We've built the page for humans, but included link structure/taxonomy to assist crawlers. So here's my question on best practices. How to handle categories with 1,000+ pages/pagination. With our most popular product categories, there might be 100,000's products in one category. My top level hub page for a category looks like www.mysite/categoryA and the page build is showing 50 products and then pagination from 1-1000+. Currently we're using rel=next for pagination and for pages like www.mysite/categoryA?page=6 we make it reference itself as canonical (not the first/top page www.mysite/categoryA). Our goal is deep crawl/indexation of our silo. I use ScreamingFrog and SEOMoz campaign crawl to sample (site takes a week+ to fully crawl) and with each of these tools it "looks" like crawlers have gotten a bit "bogged down" with large categories with tons of pagination. For example rather than crawl multiple categories or fields to get to multiple product pages, some bots will hit all 1,000 (rel=next) pages of a single category. I don't want to waste crawl budget going through 1,000 pages of a single category, versus discovering/crawling more categories. I can't seem to find a consensus as to how to approach the issue. I can't have a page that lists "all" - there's just too much, so we're going to need pagination. I'm not worried about category pagination pages cannibalizing traffic as I don't expect any (should I make pages 2-1,000) noindex and canonically reference the main/first page in the category?). Should I worry about crawlers going deep in pagination among 1 category versus getting to more top level categories? Thanks!
Moz Pro | | DrewProZ1 -
Any reason why an OSE spam Score is different for www and non-www versions of a site?
When checking the root domain links and DA in OSE pasted a non-www URL of my site and showed up with a spam score of 9. When I pasted the www. version it showed none. Does anyone think this is a cause for concern or just a bug?
Moz Pro | | insitemoz10 -
A keyword my site ranks on has recently dropped from no 2 to 51+ after weeks ranking high.
Any ideas? The full SERP analysis still puts it at no 2 so I'm quite confused
Moz Pro | | _leanneford0 -
In the hierarquical structure of the site I would like to define a campaign in a given level of that structure down to leafs. How can I define it?
I have many categories in my site and each one has others sub categorias and so on until the last categories points to products pages. I want a campaign that crawls just from a given category (or sub category) down to all produtcs pages. How can I do that? I am not sure if defining a folder would accomplish taht? Lets say I have category C, D and E. Sub categories on D are D1, D2, D3 and producs on D3 are D3P1, D3P2 and so on... How can I define a campaig on category D3 and get D3, D3P1, D3P2 crawled?
Moz Pro | | bistro0 -
In open site explorer, why are subdomain stats different to domain stats?
Hello there I have just compared some sites in Open Site Explorer. The page domain, subdomain and domain stats are all different. I get the difference between page and domain stats. But why are the subdomain stats different? There are no subdomains on the site so I expected them to be the same. What is my confused brain missing??? Many thanks Wendy
Moz Pro | | Chammy0 -
Why does Open Site Explorer show less inbound links than yahoo site Explorer?
Hello, We have a question regarding inbound link measurement. We used to measure our inbound links with yahoo site explorer. Now that it's been shut down we use opensiteexplorer.org. However, Open Site Explorer only shows a fraction of inbound links compared to yahoo site explorer. For our website www.theprintspace.co.uk yahoo site explorer measured approx. 14,000 inbound links, whereas open Site Explorer only counts approx. 3,000. This is more than 10,000 links less. For our other website www.theprintspace.de Open Site Explorer also shows 3000 links less than Yahoo. How can this be? Does Open Site Explorer count the links in a different way to Yahoo? Please explain. It would be great if you could help us with this. Thank you!
Moz Pro | | Waplington0 -
Open Site Explorer missing links
Hi, When the update of Open Site Explorer was released I noticed that the new version was missing a huge amount of links that the old version previously found. This still seems to be the case and it's pretty frustrating as we use the tool for our clients. Is this something that everybody is seeing and if so SEOMoz when do you think you'll have a solution? Many thanks
Moz Pro | | JonathanSmith0