SMX Liveblog: Advanced Technical SEO Issues
Diving into technical SEO, we have the following esteemed speakers:
- Bill Hunt, President, Back Azimuth (@billhunt)
- Maile Ohye, Sr. Developer Programs Engineer, Google (@maileohye)
- Eric Wu, VP, Growth & Product, SpinMedia (@eywu)
It’s still a little early for some and the peeps here seem to be a little slow moving this morning, but I’ve downed my cup of coffee and I’m ready to dive into technical issues with these speakers this morning. I love tackling technical SEO problems on client sites and the bigger the site the bigger the problems usually. Hopefully these experts will have a few good nuggets of information for us.
Maile Ohye: Javascript Execution and Benefits of HTTPS
“We recommend making sure Googlebot can access any embedded resource that meaningfully contributes to your site’s visible content or its layout,” Ohye said.
On May 23, Google announced they are doing more JavaScript exe. They’ve been fine tuning and were finally able to release. Shortly after they launched Fetch as Google that now shows 200 kb of a page, view text content, fix blocked resources at the rate of 500 fetches per week. You are able to select different crawlers including mobile. Use this to check and make sure Google is able to fetch all the important pieces of your site.
Modals and interstitials – they are everywhere. You need to determine if it’s worth interrupting user’s workflow. These are interrupting their workflow when it might not be necessary. Check analytics to see if it’s even needed. If you don’t want your modal or interstitial to be indexed, check the fetch as google rendering to see if it is and you can always disallow in robots.txt. Check one more time to make sure the disallow worked.
Optimize the content that is indexed. Make sure your css/js resources are crawlable. Use the fetch as Google to make sure they are rendering and remember to prioritize the solid server performance. Consider keeping (rather than deleting) old js files on the server. Sometimes Google may need these files when doing their crawling. Lastly, degrade well because not all browsers and SEs execute JS. Make sure you test.
HTTPS Benefits
Who’s prioritizing security? Security is becoming a bigger deal and there are several big sites now offering secure browsing. Why the switch
Using TLS allows for users to know they are where they expect to be on a site – authentication; it gives you data integrity and also encrypts data. TLS gives an added layer of security for the users. Google can spider https. Yes https is search friendly. Boy, that wasn’t the case a few years ago. Webmaster tools is equipped for https. You can actually verify a https site within GWT. Make sure only one version is available for crawl to avoid issues.
Crawling and indexing https sites – you want to do a 301 from http to https to avoid dup content. Serve all https resources and make sure your rel=canonical are correct. When deciding to move over to https, make sure you test your site in browsers to ensure all your resources are showing correctly.
The web is growing towards authentication, integrity and encryption so be ready. HTTPS site migration can still be search friendly if you do it right and are consistent in serving up resources.
Whew, she went through a lot of data very quickly.
Eric Wu: Ajax is like Violence – if It Doesn’t Solve your Problems, You aren’t Using It Enough
Most sites use things like jQuery, angular JavaScript or even backbone JavaScript. These help to speed up the site depending on how you utilizing them. The idea is to improve the site speed to improve the user experience and crawlability of a site. On a test, after improving the site speed Eric’s team saw an 80% increase in organic search sessions.
Google has been trying to crawl js since 2004. Over the years they’ve gotten better and better until present day where they are finally comfortable saying they can crawl it. The GWT Fetch & Render is a way to see that they’re capable of doing this now.
Eric suggests implementing Infinite Scroll. Not only because it works well on mobile but because it’s a better user experience. Use rel=next / prev when implementing Infinite Scroll.
Continuous Content uses PushState that requires a simple bit of code:
History.pushState (
{}
titleOfPage,
newURL
);
History.replaceState ();
Ajax Galleries in terms of slideshows on sites like publishers. Slideshows give you big user engagement, more social shares and numerous other benefits. Eric mentions Vox as a site that is using pushState, and rel=next / prev effectively in this case.
Defered Image Loading is something that Eric says there isn’t a good solution for…yet. Work arounds include using 1×1 blanks, skeleton screens or using low resolution as the “lazy” loading solution.
When using responsive images you can use:
• Scrset
• Polyfill
• Ua detection
In order to load images for different devices, Eric suggests using noscript.
Bill Hunt: Improving Indexibility & Relevance
As the final speaker for this session, Hunt promises to not be as “geeky” as the previous speakers.
He promises basic, so he breaks SEO down into four areas: indexability, relevance, authority, and clickability. Bill will speak about 2 of these items.
To improve indexability you need to remember that if spiders can’t get to the content they can’t store it! Improve the crawl efficiency on large sites to allow the spiders get to the content. Reduce the errors by checking them and fixing them. As development gets more complicated the more you’re going to have to tell search engines where to go and how to fetch the data you want indexed.
When submitting an XML sitemap to the search engines, check for the errors and fix any that are reported. Bing has stated that if more than 1% of pages submitted have errors, they will stop crawling the URLs in the XML sitemap. Clean up your errors folks; make things easier on the search engines. It’s not their problem to figure out, it’s yours. You don’t want to have a disconnect in the number of pages on a site and the number of pages in sitemaps. If Google and Bing bother to tell you where you have problems on your site, take notice and actually fix them.
Some common challenges that Bill has seen on sites have include:
- URL case inconsistency: site has both upper and lower case in the urls
- Page with no offers: nearly 2M soft 404 errors are due to no offers/content
- Canonical tags resulting in 2-200 duplicate pages.
Bill suggests:
- Mandating lowercase URLs
- If there are five or fewer results on a page, add noindex & nofollow
- Implement custom 404 with 404 header
- Dynamically built xml based on taxonomy logic
- Add sitemap error review to weekly workflow
Leverage hrefs on global sites. What hrefs can do is prevent duplicate content from country pages and helps the search engines understand which version is for which country and language. When doing hrefs you MUST reference the original url somewhere in the code. Many sites don’t do this and a lot of the tools don’t do this. Another mistake is referencing the incorrect country and language. Bill actually built a href builder (hrefbuilder.com) that will help you build URLs.
And with that Bill’s done. This session delivered what it promised – lots of technical SEO goodies for us geeks.
2 Replies to “SMX Liveblog: Advanced Technical SEO Issues”
This was a fantastic session. Loved how Bill brought the advanced crowd back down to the simple fundamentals.