There are a lot of different website types on the Internet. Websites differ in the way they present content and in the number of users interested in it. Bearing those differences in mind, link building strategists are aware that various websites require different approach towards link building strategies. One of the greatest sources of information for link strategists is the competitors’ benchmark.

Brief Methodology

During my tests I was using OpenSiteExplorer for Domain Authority of linking domains, link profile tool prepared by Tom Anthony (thanks for that great tool, by the way!) and CognitiveSEO (Competitive Link Analysis). I checked backlink profiles of seven website types (bank websites, blogs, discussion boards, classifieds portals, computer online stores, mobile phones operator companies and news websites). The data included in this study is exemplary. You shouldn’t rely solely on these graphs. To find out the average backlink profiles working for your industry branch, you have to conduct the research yourself. This study is just an idea. It may be necessary for you to adjust it for your purposes.

All the websites I checked were Polish and found on Google.pl. These were the top ranked websites from Google for a tested query. I chose only the websites of the tested type (websites like Wikipedia were skipped).

Key Findings

Having started up a new website, a link building strategist has to work out the best approach to link building. While preparing a link building strategy, it’s good to be aware of competitors’ backlink profiles. Different types of websites are visited by various users and as a result, they receive different backlinks.

Examples of Backlink Profiles

Below, you can see the way in which the average backlink profiles vary. There are presented the backlink profiles from seven website types, with their average values. To create the most reliable average profiles, unnatural ones were removed. In your study, you‘d better remove the unnatural backlink profiles as well – especially the ones having huge impact on the average profile curve. All graphs were created with link profile tool developed by Tom Anthony. To find out how to create them, please visit this article.

Backlink profiles of bank websites (found with a query “bank”)

Backlink profiles of information/news websites.

Information/news websites were found with a query “wiadomości” (English: "news").

Backlink profiles of blogs.

Blogs were found with a query “blog moda” (English: "fashion blog").

Backlink profiles of discussion boards.

Discussion boards were found with a query “forum młodzieżowe” (English: "youth discussion board").

As we can see from the above, youth discussion boards’ backlink profiles are quite differentiated, yet their average value isn’t very “demanding”.

Backlink profiles of classifieds websites.

Classifieds websites were found with a query “ogłoszenia” (English: "classifieds").

As we can see on the graph presented above, the backlink profiles of classifieds websites have two types. That discrepancy requires deeper analysis to find out why those profiles differ so much.

Backlink profiles of computer stores were checked on first seven websites after typing in “sklep komputerowy” query (English: "computer store"). To achieve the most natural average profile, two profiles were removed because of their unnatural appearance.

Mobile phones operators companies backlinks profiles were checked in the scope of first five results. One profile was very unnatural and so it was removed. Profile D also seems unnatural, but wasn’t removed because it has low impact on the average profile curve. Mobile phones operators companies were found with a query “telefonia komórkowa” (English: "mobile phone").

This example will be checked more thoroughly with the use of ConitiveSEO tools in the further part of this article.

What can we do with this data?

In your study you should analyze backlink profiles of all top ranking websites (including these with unnatural backlink profiles), because analysis of unnatural backlink profiles can be also valuable. Although, in order to create average backlinks profile, you should use only natural backlink profiles.

Having created the average backlinks profile, based on our competitors websites we can learn if

• our competitors have the intensive link building strategy
• our competitors are trying to gain only high quality backlinks

Owing to those information you will know, what level of backlinks quality lets you compete with the branch leaders. Knowing these answers, you’ll be able to better estimate time needed to get the organic traffic, reach high positions in search engine and successfully compete with branch leaders.

Let’s be more thorough

Let’s take a closer look at the backlinks of mobile phones operators. I’ve created a campaign in CognitiveSEO to compare backlink sources. In Competitive Link Analysis we can check what sources these websites gained links from. Thanks to this information we know which links our strategy should be focused on. That data enables counting shares (last column) of different pages with links. Backlink sources you can also check with OpenSiteExplorer.org.

Having combined the data collected from backlink profiles analysis and CognitiveSEO we have arrived at a conclusion about which QUALITY links we should collect and WHAT TYPE OF PAGES we should obtain those links from.

In the example presented above, most of our link building activities should be focused on generic websites. As we can see in the CognitiveSEO data, about 13% of backlinks in this branch comes from blogs. It may be a clue to create some offer for bloggers i.e:

• testing your products,
• high quality guest posting,
• infographic distribution.

We can also see, that ~13% of backlinks come from forums. Again, it’s a useful piece of information. Maybe it's worth starting an expert help on discussion boards? It might generate some natural links and help to promote the expert image.

Conducting such a benchmark will give you an insight into your competitors’ activities and can bring some inspiration as far as your professional activities are concerned. You have to be aware that fitting in the average backlinks profile of your branch competitors is not always the best way. First, you also have to find out if competitors’ link building strategy meets the requirements of Webmaster Guidelines. Only if it does, you can try to copy and improve their approach.

Everybody is trying to be the first in ranking, so we have to plan our link building strategy in such a way to make the backlinks profile a little bit higher quality. We can subtract ~2% from the amount of links from websites with DA 10-30, and add ~2% to the amount of links from websites with DA 40-100.

Thank to this approach, our link profile will be natural and high quality.

Conclusion

It’s hard to find a golden mean in link building strategy. While planning link building activities, we have to firstly check backlink profiles of our close competitors, and choose the ones closest to the average curve. Choosing proper activities in link building strategy is very important both for reaching high positions in Google search engine result pages and for a natural look of our backlinks profile. The awareness of profile appearance, and link sources of our competitors, will allow to create more efficient link building strategies.

I hope you enjoyed this study and that it will be useful for you while planning link building strategy.

Full methodology

Tools used in this study:

• OpenSiteExplorer
• Excel link profile tool by Tom Anthony
• CognitiveSEO

[CognitiveSEO also provides data about Domain Authority of linking websites, so if you have an account, first part of study will be faster. Remember to normalize your data.]

Presented data was normalized. All graphs are based on the percentage of Domain Authority of websites with backlinks, and not on numerical quantities of backlinks, because checked domains are unequally old, and unequally popular. Presenting percentages is more valuable and descriptive.

Backlinks were checked for seven types of websites. All the websites I checked were Polish and found on Google.pl. These were the top ranked websites from Google for a tested query. I chose only the websites of the tested type (websites like Wikipedia were skipped.).

Below is a list of queries I used for a test.

• Bank websites – found with a query “bank”
• Blogs – found with a query “blog moda” (English: "fashion blog")
• Discussion boards – found with a query “forum młodzieżowe” (English: "youth discussion board")
• Classifieds websites - found with a query “ogłoszenia” (English: "classifieds")
• Computer online stores - found with a query “sklep komputerowy” (English: "computer store")
• Mobile phones operators companies - found with a query “telefonia komórkowa” (English: "mobile phone")
• Information/news websites - found with a query “wiadomości” (English: "news")