Developer, Meet SEO
I have a lot of favorite things about my job but one of my most favorite things is that I am constantly learning. When I learn something that is totally new to me — a new concept, approach or technology perhaps — I find it exhilarating and satisfying. I had that feeling years ago when I broached the basics of SEO. When I realized how search engines think it blew my mind that you can cater a website to be favored in search results. I learned quite a bit about the basics of SEO.
But like with all technology, what I knew about SEO from several years back all amounted to being either obsolete or incomplete. I figured a good page title, human-readable URL and a handful of targeted keywords on a page would get a site down the road with SEO but I learned recently that this is very much not the case, especially if you’re working in a competitive niche.
Learning Things the Hard Way
One of my least favorite things about this job is that I’ve occasionally had to learn things the hard way (read: “from my mistakes”). At TopFlightApps we work hard to add quality and value to every level of our client’s mobile app or website. If we just build your site and send you on your way then we’re doing you a disservice so we are always working on ways to help ensure the success rates of our clients.
We have taken to focusing on SEO as part of our active improvements and our client TheIndex was an early target of our SEO analysis and boy did it need it!
TheIndex (mysurgicalindex.com) is a multi-vendor marketplace where users can buy and sell their surplus medical supplies. It was built with Magento with a theme and several extensions.
So when it came to my attention that the home page’s title said “SM Market — Layout 2” and ranked second for the search of “site:mysurgicalindex.com” you can probably imagine the rush of blood to my cheeks and sweat to my palms. If there’s one thing I know about SEO it is the basics! That was just one of thousands of issues across the many pages of the site.
The Learning Experience
We performed a crawl on Moz.com and it found, at worst, 29k+ meta-information issues on the site. At the time of writing, there are about 8000 products on the site so the issues compounded pretty quickly. Our SEO team gave us a video with further analysis and I hit the ground running getting the changes in place.
Robots.txt is Critical
Every page was being indexed by Google because nothing was telling it not to and I wasn’t really aware of how big a problem this can be for SEO. I adjusted the file so that Google will not index search result pages, unnecessary URL parameters along with several Magento-specific pages unfit for search results.
Every website is different but if you’re working with a CMS like Magento or Wordpress you can find awesome starting points online. That is where I started and it was an almost immediate help.
Duplicate Content Issues & “rel=canonical”
This was entirely new to me. When Google looks at your site it can see several versions of it. You might have https://mysite.com or http://mysite.com or http://www.mysite.com or https://www.mysite.com or literally any other combination of HTTP(S) and subdomains.
If you load a page and don’t have a meta tag defining the canonical, meaning the “preferred”, version of a website then Google won’t know which is which and treat them all as completely different pages. If you have 4+ pages on your site with the same content Google will think you’re trying to pull a fast one and dock your site.
So if you have a canonical tag saying that https://mysite.com is your preferred location for all of the above URLs, then Google won’t penalize you for having 4 (or however many) versions of a page of your site.
This was another compounded issue that Moz.com found for us. I was upset to see so many errors at first but with hundreds of pages having this same error it was a huge fix for TheIndex’s SEO.
Check out Yoast’s post about rel=canonical for more information.
Duplicate and Low Quality Headings and URLs
TheIndex allows vendors to upload their own products so the owner doesn’t really have full direct control over the quality or length of titles entered by vendors or the descriptions of products. This works somewhat similarly to eBay. There are products in the system that have the SKU as the name and the description. For the purposes of TheIndex, this is fine since someone looking for an SKU will find it, but for SEO this is not recommended.
So this had to be mitigated with TheIndex. Fortunately, we found a wonderful extension that gives the power to create templates for products, categories, and filtered navigation pages. Titles are now a combination of the product manufacturer and SKU and its meta information is laden with the product’s description, SKU, and manufacturer.
The URLs were often times long or otherwise unreadable by a human. They gave little information about what might lie on the other side and really looked like a string of letters and numbers. This cannot always be changed, but using the manufacturer and SKU instead has reduced the number of bad URLs by and by including more crucial and dynamic product information in the titles we lopped of another few thousand of Moz.com’s errors.
Lessons Well Learnt
This article is just a taste of what I learned as a developer working on SEO. I’m not going to become specialized in SEO or start proselytizing for it but I think it should be something every web developer has a solid understanding of. If you’re building something for the web you should have a holistic view of how your final product will appear to the world when Google’s spiders start crawling its nooks and crannies and report back to your potential users’ search results.
Treating SEO as a first-class web development task is now the rule over the exception for me. It is something that ought to be done in parallel with the rest of a project and considered as early as the design phase. I’m really happy to have updated my SEO knowledge and I hope that this might encourage other developers to reconsider what they think they know about it.
