We Need a Real Mobile SEO Methodology

A maddening issue plagues the SEO community, a stone in our collective shoe that we keep trying to ignore, but that at least once a year rolls right back under the most sensitive part of our foot. We keep walking because the fast pace of the industry and the day-to-day needs of our clients make it so we don’t have time to stop, untie the laces, and pop the shoe off to shake the stone out. But sooner or later we’re going to have to, come what may.

I’m speaking of the disconnect between the primacy of mobile devices in search — both in Google’s eyes and users’ — and the desktop-exclusive perspective that we professionals remain locked into. Daily Google search queries on mobile outpaced desktop years ago, and the Google Speed algorithm update promised for July of this year, in which mobile page speed becomes an official ranking factor, constitutes Google’s third big effort to jam home the same mobile-first message in as many years (following the “Mobile-Friendly Search Results” update in 2015 and the “Mobile-First Index” update of 2016, and heck, you might even count AMP here too). But SEO practitioners simply cannot adopt a felt, experienced appreciation for this changed landscape, for the simple reason that desktop computers are still, overwhelmingly, the devices that we use to do the work.

I did a complete SEO audit of a new client’s site a couple of weeks ago — my first in a while — and was newly struck by how little mobile search and mobile user experience factor into the standard audit slate, despite our constant emphasis of mobile in conversation, and by the fact that the work that is done to address these factors is mostly done with simulators. I click the iPhone viewport emulator in Chrome DevTools and think I’m experiencing the mobile site. But am I really, if I don’t have to use my fingers to tap and scroll? I conduct a 3G fetch of a page using WebPagetest and think I understand how quickly the site will load on a smartphone over a 3G network, because I’m given a measure in milliseconds. But do I really understand, as a user would, if I never experience the page load wait firsthand? I use SEMrush to retrieve mobile rankings for my clients alongside desktop rankings. But does that mean I know how my clients’ sites are actually situated in mobile search results relative to their competitors, and how different the SERP experiences are from those of desktop, regardless of precise ranking position?

This is not a call to harvest more data. It’s a call to close the gap of feeling between the mobile-first SEO world that we are told is the real, meaningful one, and the desktop-first SEO world that we actually do the work in. This means we have to find a way to rely less on numbers and emulators telling us what our sites are like on mobile, and actually get into the trenches, which means starting to do at least some of the work itself on mobile devices.

The problem with that is that I have no idea how to do it.

Traditionally, SEO audits — and let’s just start there, for the purposes of this post — rely on:

  1. Data about markup and content, returned by crawlers (e.g. Screaming Frog),
  2. Data about site performance, authority, and search engine accessibility, returned by browser-based tools that we trust innately because they are made by Google/Bing/Moz/SEMrush, etc. (or self-directed headless browsing, for the more adventurous among us), and
  3. Firsthand observations that we make about the site by browsing it.

There are cases in which device type will affect the findings in Categories 1 & 2, but when we’re talking about bringing the mobile experience to bear on an SEO audit, we’re mostly, necessarily, going to be talking about the stuff in Category 3.

And even the stuff in Category 3 can be further split into two subcategories, which I might describe thusly:

  1. Stuff that simply is either there or not there, e.g. a robots.txt file, and
  2. Stuff that is fundamentally about the site experience.

And so I think the work we need to start doing is largely going to take the form of additions to Subcategory 3b, i.e. we’re going to have to carve out space for more and different observations. Mobile-specific ones.

Here are some I can think of, off the top of my head:

  • Disabling JavaScript in your mobile browser and visiting the most important pages — i.e. the most popular search landing pages, and the pages most critical to the site’s funnel — over the cell network, to make note of what is broken and what takes too long to load;
  • Checking the mobile navigation with JavaScript disabled to make sure it works and is comfortable to use;
  • Walking through the various conversion funnels on the mobile site as well as the desktop site to make sure that the mobile experience isn’t dreadful by comparison;
  • If the client is using AMP, making sure that AMP versions of key pages are being returned in search at a decent ranking and with proper metadata;
  • Manually, experientially checking the mobile site over the cell network to see how quickly images load compared to desktop;
  • Manually conducting mobile searches on the client’s brand name and on major non-branded keywords, and screenshotting your findings to hold up against the desktop equivalents and determine how the experiences differ;
  • (We’re seeing this less and less, thankfully, but:) checking to be sure that if the mobile site loads at distinct URLs and/or uses dynamic serving, that all the proper precautions have been taken in the HTML to keep the site optimized for search despite the handicap.

This testing would, of course, require access to multiple devices. One can always use one’s own phone, which most of us have, but at least in the interest of a caveat I must of course mention that the principle of maximum rigor would have us test all of the above on both an iPhone and an Android phone, at the very least, and would also have us work hard to keep pace with new phones and mobile OSes as they evolve and roll out. Then after we open that door, we find ourselves facing more: one that leads to tablet testing, one that leads to testing deliberately on older, near-obsolete phones, etc.

For the purposes of simply introducing the idea here and now, I’m going to keep those doors shut. I need a starting place, and having given at least this much thought to the matter, here’s where I’ve decided the starting place is: the next time I have a new SEO audit to do, I’m going to add the above bullets to my task list, see what I discover by doing so, and then anonymize and report my discoveries in a future post. So go ahead and consider this the first post in a series.

And in the meantime, tell me: what else should I be looking for? What are some other mobile experience factors that should be considered as part of a modern SEO audit? What else do we stand to gain by devoting more time to poking around our clients’ sites on our phones?

Alternatively, am I missing something, or getting something wrong? Is this not worthwhile after all? If not, why not?

Related Posts

Leave a Reply

Your email address will not be published.