Last week a Dutch company specialized in SEO tools called Lipperhey announced a new link index. Within 12 months they want to become the largest publicly accessible link database in the world. Together with Target Holding a part of the University of Groningen and 1,000 volunteers they started this project.
The goal of Lipperhey is to map 80% of the links in the internet. It’s a bold statement and hard to verify but informing about this statement they respond:
“When you start project of this size you have to formulate goals as a company. The first questions that arises is: What is a link and do we count internal links, as other tools do? This is an essential question. Because including internal links increases the amount of links 35 times. If we continue to expand the way we do at the moment, we will be indexing 1 billion unique external links a day.
A good guide for concretizing the goal is the “Industry Brief” by Verisign. When we take that as a starting point, there would be 209.8 million registered domains at the moment, from which ten country TLDs (.de, .uk, .nl, .eu, .cn, .ru, .ar, .it, .pl) and the top levels .com, .org, .net and .info together form 80% of all web pages in the world. Our first calculations show that 80% of the links will be somewhere near 85 billion unique external links, and growing.”
They want to distinguish themselves from the competition by incorporating artificial intelligence into the tool. “We are the only tool that links science to link analysis”. This way they claim to be able to determine the sentiment and context of links by looking at surrounding information of the link (this will be included in later releases). A new way of looking at link analysis indeed. But will it work? The success of the tool will be dependent on the usage of the data. Will it be actionable? Will people see the benefits of the extra information presented?
The Lipperhey Link Explorer has a user interface that reminds me of Open Site Explorer. It’s a clean interface with only the necessary items visible. The landing page shows me a text box where I can add my URL to explore. There’s also a possibility to add a URL to compare the link data with. Below we see a number indicating the links indexed by Lipperhey at a moment. Hovering over this number we see the statement “Yes, this is real time”. According to Lipperhey it’s real time with a margin of 60 minutes.
Next there is a link to “Become part of the project”. Following this link you can install a Spider to help Lipperhey index all the website of the world. This is bit of an odd statement realizing they index links, not websites. They’ve got spidering tools for Windows, Mac OS X and (Ubuntu) Linux and a short FAQ handling questions considering purpose, effects and configurations.
My first attempt to get data for www.stateofdigital.com shows a summary with the following stats: 1,037 incoming links and 27 referring websites. The tool couldn’t give a number of outgoing links yet (not indexed yet). The average link value is showing 2/10.
There are four tabs with extended information on some metrics. The tab incoming links shows the links pointing to the websites with the page title and url, type of link (text/image), description (anchor text) and authority. Although the tool apparently found 1,037 it only shows data of 26 links. What’s interesting is that it only shows links from Netherlands based websites. Apparently the indexing started in The Netherlands. It is not possible to sort or filter the links. The authority mentioned in this tab has an information button explaining why authority is important while I would expect an explanation of how it is calculated.
The second tab shows a list of referring websites. I would expect a list of domains here but it again shows a list of URLs with page titles. Only in this tab there’s only one item per website and in the columns it shows the number of links (probably for the root domain), the date added (I don’t know for which link, possibly the first found) and again the authority of the referring website.
The third tab shows the anchor text distribution. Here we see the anchor texts used in the back links with the number of anchor text for each unique text and the percentage of the total of back links with that specific text. Nothing special here.
The last tab shows a list of link metrics for each link. Here we see the usual suspects: whether it’s nofollowed by either, rel=”nofollow” or a META nofollow and whether it’s a redirected link. Next to these metrics we see in other tools they also show the link value of each link. According to Lipperhey a higher value represents a higher clickthrough probability. The formula for the link value is the following:
It’s a nice way to show the value of a specific link, but there are also many arguments why it’s not a solid metric. For example, it only shows the value of the link within the specific website, but not relative to other websites. A link on a deeper page crowded by more links on a high traffic website could be more valuable than a link on a homepage of a low traffic website.
The tool also offers a CSV export of the report. This export shows all the links from link metrics tab. For these links the following metrics are shown:
- Meta robots nofollow
- Image SRC
- Image ALT
- Link Count
- Link Value
However some metrics don’t match exactly with the data from the tabs in the report. It’s seems the export is now still limited to a maximum of 500 links, showing the links with the highest authority.
When comparing two URLs the report compares the number of incoming links, referring websites and the top 10 most used anchor texts. It’s a nice short comparison but doesn’t really provide actionable insights. This would be a nice place to show where possible improvements could be made.
To check how many links Lipperhey has found at the moment I did some number crunching. In the table below I compared the number of links for a few large international sites and a few Dutch sites in different free tools. What’s interesting that when looking at Dutch sites Lipperhey is already quite competitive based on pure numbers. Considering they’ve probably started crawling in The Netherlands, this looks promising for the future.
|Lipperhey||Open site explorer||Yahoo! Site explorer||Majestic Site Explorer|
So, now it´s time for a judgement. How do I rate this tool?
First of all I like the user interface. It works for me. Also, I like the thought of adding new kinds of data, although they don’t always seem solid to me. However we have to wait for new metrics like sentiment and context which will be included in later releases.
Besides these new kinds of data there are still a few things missing in this tool. I’m missing sorting and filtering options for the data presented. Furthermore I’m missing specific kinds of data. For example: OSE shows page authority/domain authority, Lipperhey shows average link value. But what does average value say? Google.com has an average link value of 4/10. Is that bad? It’s not actionable.
Finally I think it’s still missing clear added value. If want to create a large user base they have add improvements to this tool. The start looks promising, but it’s not ready yet to be a good alternative for the major paid (or partly free) tools from the big guys. The planned developments will show how well they thought it all through. The main reason to use it at the moment will be because it’s free. For dutch websites it’s already a good alternative for other tools. For international websites, not (yet). I will be following the developments closely, it has potential.