Skip to main content

Mobile

Two Exercises for a Simple, Real-Life Mobile SEO Audit

Mercedes made headlines recently with their revamped mobile site. The good news is that they were able to increase mobile traffic 85% year to date, and 170% over last year. The bad news is that when I looked at their revamped site it was evident that they didn’t account for SEO as part of the redesign, and could have driven mobile traffic up much more if they had.  Recently in Search Engine Land I explained that mobile SEO is not a myth. To further prove that it exists I’m going to go through a basic mobile SEO audit for Mercedes’ new site, to demonstrate how one brand failed to take advantage of mobile search traffic by thinking about how mobility affects search behavior and site architecture. Hopefully, this exercise will help the rest of you avoid the same mistakes.

Basic [brand + “mobile”] Search in Mobile

When I audit a mobile site, one of the first things that I’ll do is search on the phrase “[insert brand name here] mobile site” in order to see if a brand can be found for navigational mobile queries. In this case we’ll use “Mercedes mobile site”, which according to the Google Adwords Keyword tool gets about 1,300 searches per month in Google.
Entering these keywords in Google should return the m.mbusa.com site, since the query is navigational and there’s little competition. However, when I entered the query, no such website was found.
In fact, the first result was Mercedes.mobi, which was the only thing that looked like an official site on the page. The rest of the articles had to do with the recent site redesign.
Doing a site: search in Google, it became evident that the site was not listed for the navigational query because it was not eligible for that query—that the site developers had neutered it by nofollowing the site with robots.txt.


Here is a look at the Robots.txt file for the site:

Two clear, and common, problems become evident:
A) Multiple sites competing for the same keywords
A site appears for the navigational keyword, but it’s not the site they just redesigned. It’s very likely that Mercedes would prefer that their best performing, most relevant site appears in search results, and Mercedes.mobi likely isn’t it.
To remedy this, some brand manager at Mercedes would have to decide which mobile site she wants to appear, and that site would then be optimized for search. If it’s not Mercedes.mobi, then we would focus our optimization efforts on m.mbusa.com
To be fair, this problem exists with the desktop site as well, as both mbusa.com and Mercedes-benz.com/en are often competing for the same keywords, even if Mercedes-Benz.com and mbusa.com theoretically have different audiences. This is a mobile-specific problem, however, because Mercedes also has this dotmobi indexed, which compounds the problem with the mobile-specific domain.
B) One site excluded from search entirely
The most pressing problem is the exclusion with robots.txt, which happens more often with mobile sites than you might think. Often the developers or well-meaning but uninformed SEOs will exclude the site with robots.txt in order to ensure that the site is not seen as duplicate content by Google. However, this is unnecessary, as I’ve been saying for a few years now, and as Google has confirmed in their recent mobile search guidelines. If mobile URLs are used, it’s only necessary to add switchboard tags to the mobile and desktop sites to let Google know which site is preferred for mobile searchers. If you’re not familiar with switchboard tags, Google explained them in their recent smartphone guidelines:
1. Annotation in the HTML
On the desktop page, add:

 

and on the mobile page, the required annotation should be:

 

This rel=”canonical” tag on the mobile URL pointing to the desktop page is required.
Using this method, sites can ensure Google shows the mobile site to mobile searchers and the desktop site to desktop users, and not have to worry about split link equity.
As it is the m.mbusa.com site is a non-sequitur when it comes to Google mobile search traffic, as the only hope it has for getting traffic from Google is redirecting from the desktop URL. Adding the switchboard tags removes the redirect, puts the m.mbusa.com URL in search results (which might give them a slight boost in CTR), and improves the user experience from search overall.
2. Align Site Architecture with Mobile Search Behavior
The next exercise is simple: comparing mobile and desktop searches for the brand and nonbrand terms, and comparing those to the site architecture of the site being audited.
All of this is moot for Mercedes since the entire mobile site has been nofollowed, but we can still do the exercise as a hypothetical, for when the indexing problem is fixed.
The first thing that we want to do is look at their brand keywords to ensure the concepts people are looking for on mobile devices related to the brand are prominent (or at least represented somewhere) on the site.
For this you can use the Google keyword tool, and enter just the brand name.

Download both the “All mobile devices” report and the “Desktops and laptops” reports. Use Excel’s Vlookup function to map the information into a report that looks like this:

Note that for this example I’ve only pulled U.S. search volume for mobile and desktop keywords with over 1,000 searches each but for your purposes the long tail may be valuable.
Once we have this report we can look for keywords that both have a lot of volume from mobile devices and have more than 30% of the total available search volume between mobile and desktop (represented in green). A couple of things stand out for me when scanning the list:
A) There are a few concepts with a lot of volume from mobile devices that aren’t represented on m.mbusa.com. For example, Mercedes AMG has a lot of search volume from mobile devices, and a lot relative to the total; but the site is on a separate domain: Mercedes-amg.com. There is a mobile version of that site, but it might as well be a desktop site considering how it looks when the phone is oriented vertically.

What’s more, there is a page on m.mbusa.com devoted to this model of Mercedes, but it doesn’t resolve.

Not supported yet? Or not still supported? Either way, it’s a bad user experience and something that is unlikely to rank for these contextually relevant, high volume terms.
AMG isn’t the only concept for which there is search volume but no content. The most egregious is probably the car types, represented by the keyword [Mercedes suv]. There are no category pages representing all of the SUVs that Mercedes has, nor coupes, sports cars, etc. This is because the site is using a transcoder that crawls a site and builds reformatted versions of the pages automatically, and they can’t reformat a page that doesn’t exist.
If you’re not familiar with transcoders, many brands choose mobile solutions that transcode desktop pages because they don’t require internal resources or a lot of budget to implement, but they can have many disadvantages when it comes to SEO. The biggest disadvantage with most is that they only transcode desktop pages and can’t add pages that make sense for the mobile paradigm, or pages that don’t exist on the desktop site.
Because the car type pages on the desktop site use hashtags in the URLs they are inaccessible to spiders, and can’t be indexed as individual pages. There are workarounds, like SWF address, for making these URLs accessible with this technology, or Mercedes could choose to redesign the site with static URLs.

Either way they’re missing out on a big opportunity, as our initial research has shown that those in the market for a new vehicle who don’t yet know what brand they’re looking for mostly search by car type. We categorized the non-brand keywords and put them in a pivot table to get a better sense of the opportunity available, and discovered that the lion’s share of the search volume comes from car types.

There are also many searches for mobile wallpaper, which Mercedes has on their Mercedes.mobi site but not on m.mbusa.com.

When we look at the search volume for mobile wallpaper, and mobile wallpaper for cars, it’s clearly the one concept that is relevant to the brand that searchers are looking for on mobile devices much more than desktops; but it’s not included at all on this site.
It’s difficult to be visible for a query if you don’t have content that’s relevant to it, so for this client, we would recommend building out specific pages for relevant car types and characteristics of those car types, as well as combining the Mercedes.mobi site with the m.mbusa.com site (including the wallpaper) in order to better align their site’s information architecture with what consumers are actually looking for.
B) The second part of this that struck me as odd is that when we look at local search terms the volume doesn’t appear to support such prominent placement on the homepage.
When we look at the home page we see that the primary navigation consists of three calls to action: select a vehicle, find a dealer and special offers.

Now it could be that these were selected for reasons other than search behavior, but from the keywords that we’ve seen local information is not as important to this audience as it is to mobile searchers in general. In fact, if we look at the branded keywords above, the few local keywords that appear (Mercedes Benz Houston, Fletcher Jones Mercedes, Mercedes Benz Chicago, Mercedes Benz of Buckhead, Mercedes Benz Dallas, etc.) actually have less search volume from mobile devices than the average. I don’t know if I would remove it from the homepage, but at least make car types more prominent, as they’re currently absent from the site.

The Results

In performing these two exercises we can see clearly that Mercedes needs help on the mobile SEO front. Based on the initial audit we would recommend the following (in order of priority):

  1. Allow Google to crawl m.mbusa.com by changing robots.txt file of the site

  2. Implement switchboard tags in the short term to allow Google to understand the relationship between mobile and desktop pages

  3. Consolidate duplicate content on Mercedes.mobi and m.mbusa.com with canonical tags or redirects.

  4. Address information architecture of the desktop site, as it’s preventing traffic from the desktop and mobile site by excluding certain pages like category pages for car types and characteristics.

  5. Address microsites that are duplicating content both for mobile and desktop properties.

  6. In the long term we would recommend either using responsive web design for one site, or ideally a hybrid of responsive and mobile-only pages so that we can continue to offer things like mobile wallpaper that are important for loyalty marketing and branding. The search behavior is different enough to warrant a few mobile pages, but probably not an entire site.

If this were an actual mobile SEO audit it would be much longer and more in-depth, but I wanted to present a limited version here for this reason:
Mobility changes the SEO game. Whether you have a mobile or responsive site there are issues with search behavior, site architecture and link building that SEOs need to address now if they are going to do their jobs fully in a world where one out of every seven people on the planet own a smartphone. This is one example of how I might do it for a client. Hopefully, you have your own ideas and we can have a discussion that will advance the practice, as we’ve done with traditional SEO for more than ten years. But the last thing we can do is ignore it, because these problems aren’t going to detect and resolve themselves.
Hopefully, this simple audit helps you do actual mobile SEO on your own sites. If you have questions, feel free to ask in the comments, or on my blog.
Disclosure: Resolution Media does some paid search for Mercedes Benz, but has not been engaged for SEO.

Perficient Digital Author

More from this Author

Categories
Follow Us