Some folks in our organization feel that our SEO is actually hurt (and not helped) by a hosted Community, that we should instead have a community that is fully integrated into our web site. How can I convince them otherwise?
Solved! Go to Solution.
Hi Keith:
I'm so pleased you brought this up. Part of my past involved a large search engine company that started with "G" and ends in "e", and as such, I am well-versed in SEO, so I thought I'd respond. (Search "Patrick Riley, Google" if you want to know more.)
In terms of overall SEO, there is no difference to search engines if you have a hosted or in-house community. However, the position of the domain name can make a different, which is why Lithium recommends a subdomain (ex: community.domainname.com). This is significantly better than lower down in the domain name (ex: www.domainname.com/support/community).
Google will start to cache and add your community to the overall SEO of your domain name, and we’ve seen communities make a significant bump in Google’s ranking (from PageRank 4 to PageRank 7), which is *huge* considering the max is 10, and the scale is a log formula.
However, there are some reasons why hosted communities are actually better than in-house software communities:
1. Speed - As Google states, it’s important for its spiders to be able to access as many pages in your sitemap as possible. Speed is a huge factor in how many pages are indexed by Google (and others). Lithium has engineers constantly improving the performance of how fast our pages load, and have the industry’s fastest
Let’s juxtapose two similar Lithium (hosted) and Jive (software in-house) communities (I’m using a web testing service… many are available to try: http://www.selfseo.com/website_speed_test.php)
You’ll notice they both use the subdomain to show their community, and are similar-sized companies in terms of traffic, visitors and conversations.
Lithium Hosted Community (Norton)
community.norton.com/norton/board/message?board.id
155.06 KB 1.13 seconds 0.01 seconds
Jive Software In-house (Intel)
communities.intel.com/thread/2786?tstart=0
107.77 KB 2.33 seconds 0.02 seconds
Ouch... the hosted product is finished with its "victory lap" before the in-house software finishes. Despite the Lithium-powered community page being significantly larger (longer thread), the hosted Lithium community loaded more than *twice*as fast as Intel’s in-house community software. This also effects how many times Google will come and index the communities (prefers faster pages), so likely the Lithium-powered community will have more real-time updates in Google’s search results.
This is a significant different both for users visiting the pages, and especially for SEO (Google only has so much time to scan as many pages as possible.) This is even stated on Google’s own blogs.
As a result, let’s look at how many pages Google cache’s in these two similar high tech communities:
Lithium Hosted Communtiy
Norton: Google scans and indexes 145,000 pages.
http://www.google.com/#hl=en&q=site%3Acommunity.no
Jive Software In-house
Intel: Google scans and indexes 58,500 pages.
This is not the total number of pages of the communities, but rather, the only ones that Google spiders can get to under a certain period of time. Google (and Bing, Yahoo, etc) indexes 3 times as many pages with the Lithium hosted community verses the software community. This also means Lithium helps the Norton.com domain roughly 3 times as much in terms of traffic and PageRank as does the Jive script In-house product.
2. Guaranteed uptime. Lithium guarantees the industry’s highest uptime of your community. In-house software doesn’t give you any guarantees. It’s up to you to keep your community up. If your community is ever down, Google will be unable to index your site, and make a note of your site’s performance. If your site is found to be inaccessible, the frequency of Google spidering will decrease.
3. Scalability and Load Balancing. Lithium’s hosted service offers instantly scalability and load balancing. If you get many hits to your community at one time (ex: a popular support issue), Lithium’s hosted offering will be able to take intense amounts of traffic. A great real-life example of this was when Linksys has some driver issues, and received a barrage of traffic… the Lithium community stood up strong and offered unparrelled performance.
4. SEO-optimized service. Lithium’s hosted communities are highly SEO’d, and an average Lithium community gets about 70-80% of the visitor traffic from Google searches. Over 5 billion conversations are viewed per month with Lithium’s hosted product. No one else can say that.
Honestly, I could go on and on about why hosted is better . Feel free to ask an SEO consultant your question, and they will return the same or similar recommendations.
Awesome question. Please feel free to email me at patrick.riley at lithium if you want any more information on how to optimize your Lithium-powered community so you can be running your own "victory laps." ;-)
That was an excellent explanation - even I think I understood it, and I know nothing about SEO!
Jane
This was a great explanation of how to set up a forum. I think to many webmasters on't worry about the affects just the time it takes. Thanks
Hi Patrick,
I was just checking out the speed test and number of pages indexed for my own community, and now I see that the two communities you referenced are much more similar than they were when you first posted this. For example, I get Norton 2.11 seconds and 237,000 pages and Intel 0.88 seconds and 165,000 pages. Do you know what drove this change?
Our community comes in with similar numbers, 1.59 seconds and 243,000 pages.
Thanks,
Laura
Hi Laura:
Thanks for your question. The test results will vary if you just run one test, potentially due to traffic at that time, or other activity going on in the respective communities (incremental indexing, peak traffic of visitors, etc.) My original numbers were based on an average of 20 tests.
I have run these tests hundreds of times, and have never seen Intel's (Jive) discussion thread page with similar # of comments load faster than Norton's (Lithium). Here are three from today run at 9am, 10am, and 11am today (attached). I also checked your site (forums.ni.com) vs. a similar Intel forum page and notice your Lithium-powered community is on average more than twice as fast as the Jive equivalent (both have 13 comments, similar size of page, etc.)
communities.intel.com/thread/2786?tstart=0
http://forums.ni.com/ni/board/message?board.id=180
Hi Patrick,
I was thinking about this a bit more and had another question. Do you have an article where this comment is discussed in more detail? We use a combination of community/domainname.com and domainname.com/community and I don't know that we see a difference in page rank. Thanks!
PatrickR wrote:In terms of overall SEO, there is no difference to search engines if you have a hosted or in-house community. However, the position of the domain name can make a different, which is why Lithium recommends a subdomain (ex: community.domainname.com). This is significantly better than lower down in the domain name (ex: www.domainname.com/support/community).
Google will start to cache and add your community to the overall SEO of your domain name, and we’ve seen communities make a significant bump in Google’s ranking (from PageRank 4 to PageRank 7), which is *huge* considering the max is 10, and the scale is a log formula.
http://www.seomoz.org/blog/seo-cheat-sheet-anatomy
Clearly demonstrates that Google gives higher priority to subdomain than folder or path/page. You'll also note this article was just composed last month using **LOTS** of SEO data. It's hard to tell from a user perspective the difference of posted PageRank (because it's just a log formula of 0-9), but I assure you, a PR of 5 and another PR of 5 could have huge differences in page optimization.
Best regards,
patrick