Don’t follow the leader: Avoid these 5 common e-commerce SEO mistakes
We often look to industry leaders for inspiration, but columnist Brian Weiss notes that big e-commerce players are often making critical SEO mistakes that should not be emulated.
Competitive research is an important part of any SEO program — after all, it’s a zero-sum game that we’re playing. However, there is often a tendency for companies to become fixated on what dominant competitors in the marketplace are doing. The assumption is that because they’re getting the most SEO traffic, they must be doing things right.
In many industries, it is true that the high SEO traffic sites really are doing an exceptional job. But in the world of e-commerce, this is often not the case. Many of the highest traffic e-commerce sites are doing things that are objectively bad for SEO. It turns out that a strong backlink profile and other prominent brand signals can make up for an awful lot of mistakes.
Getting things right for enterprise e-commerce SEO can be really challenging. You often have to merge very different sources of product data into a single system and make everything work. There are more pages than you could ever curate manually. And in most cases, SEO is not the largest driver of traffic and may have to take a back seat to other priorities. It’s tough.
Eventually, people are going to figure out how to address the issues that make e-commerce SEO so cumbersome and hard to scale. Sites that apply these new techniques will gain an advantage, and then everyone will race to copy them and this article will be outdated. I believe that point is still some years away.
Until then, there are opportunities to gain an SEO advantage over most of the major e-commerce players by simply avoiding their most common mistakes.
1. Faceted navigation disasters
When faceted navigation isn’t controlled, you can often end up with more category URLs, by orders of magnitude, than total products on the site. Clearly, something is wrong with that picture.
On the other end of the spectrum, you have companies that are so scared of creating too many pages that they noindex their entire faceted navigation or canonical everything to the root page. Doing this can prevent indexation of potentially valuable pages (usually ones with two or one attributes selected) and it still may not fix the crawl problems that their navigation poses.
There is a middle path, and few try to walk it. While fixing your filtered navigation is an entire topic of its own, a good starting point is to consider using dynamic AJAX sorting for thin attributes, so users can refine the product set without changing the URL.
2. Slow site speed
There is plenty of readily available data about the impact of site speed on conversion and bounce rates. A couple of seconds can make an enormous difference in user engagement. So why do retailers seem to be competing to load the most external scripts? The retail market is underinvested in speed and overinvested in lag-inducing features that often have marginal benefits and may even serve to overwhelm the user.
My experience is that the SEO benefits of page speed are not yet as substantial as the conversion optimization impact. With all the information Google is sharing about the user benefits of fast, streamlined sites, it’s only a matter of time until speed becomes a more prominent ranking factor. However, when UX impact is also taken into account, there’s no reason to wait.
3. Reliance on XML sitemaps for indexation
If there is one simple piece of SEO wisdom that every enterprise manager should remember, it’s that each page needs to have a crawl path to have a chance to rank for competitive queries. There are many unique and exciting ways (from the perspective of someone who is paid to fix websites) that sites are able to orphan a large percentage of their product or other important pages from their browsable architecture.
Possibilities include broken pagination, creating nearly infinite URL spaces, and any form of link generation logic that doesn’t systematically ensure that every product has a crawl path.
If you’re unsure about whether you have an adequate crawl path, crawl your site and see if all your important pages are showing up. If you are not able to do a complete crawl of your site, that means either that you have too many pages or you need a better crawler. If you have a very large site, you likely need help with both. And if you’re spending lots of time looking at the sitemaps dashboard in Google Search Console, wondering why your pages aren’t being indexed, it’s most likely because they don’t have a good crawl path.
4. Using tags completely wrong
Many e-commerce sites have conflicting tagging signals on their category pages and tagging structures that are suboptimal. I have seen at least two Fortune 500 owned e-commerce sites that were making all the pages on their site canonical to the home page, which is equivalent to telling Google that none of the other pages on the site have anything else to offer. I have seen more sites than I can count on one hand do their pagination tagging incorrectly, which is surprising, because it’s a plainly spelled-out specification.
I suspect that Google’s assumed omniscience sometimes hinders the careful adoption of standards. People think they can get it close enough and Google will figure it out. Sometimes they do. Sometimes they don’t. And sometimes, even if Google can figure out all your mistakes, it’s still a loss — especially if they are having to crawl extra pages to do so.
5. Ugly URLs
Here’s a thought experiment. Let’s set SEO aside for a moment and look at two different URLs that we might see in a SERP:
Site 1: www.madfancylad.com/c/armani-fedoras
Site 2: www.bromendous.com/search?product%20line=fedora&brand=Armani&REFID=23ghaWHY23093482
Which site seems more likely to make things easy for their shoppers, and which site seems more likely to make things easy for themselves? What kind of conscious and unconscious assumptions might a shopper make about each?
My experience is that short, clear and concise URLs tend to rank well and get more traffic than long, parameter-laden addresses. There are some correlational studies that support this observation. I don’t consider any of them definitive — but I know what I would choose to do for my site.
Opinions expressed in this article are those of the guest author and not necessarily MarTech. Staff authors are listed here.