What’s Your AMP Traffic Really Doing? Set Up Reporting in 10 Minutes

Posted by Jeremy_Gottlieb

The other day, my colleague Tom Capper wrote a post about getting more traffic when you can’t rank any higher. I was really pleased that he wrote it, because it tackles a challenge I think about all the time. As SEOs, our hands are tied: we’re often not able to make product-level decisions that could create new markets, and we’re not Google’s algorithms — we can’t force a particular page to rank higher. What’s an SEO to do?

What if we shifted focus from transactional queries (for e-commerce, B2C, or B2B sites) and focused on the informational type of queries that are one, two, three, and possibly four or more interactions away from actually yielding a conversion? These types of queries are often quite conversational (i.e. “what are the best bodyweight workouts?”) and very well could lead to conversions down the road if you’re try to sell something (like fitness-related products or supplements).

If we shift our focus to queries like the question I just posed, could we potentially enter more niches for search and open up more traffic? I’d hypothesize yes — and for some, driving this additional traffic is all one needs; whatever happens with that traffic is irrelevant. Personally, I’d rather drive qualified, relevant traffic to a client and then figure out how we can monetize that traffic down the road.

To accomplish this, over the past year I’ve been thinking a lot about Accelerated Mobile Pages (AMP).


What are Accelerated Mobile Pages?

According to Google,

“The AMP Project is an open-source initiative aiming to make the web better for all. The project enables the creation of websites and ads that are consistently fast, beautiful, and high-performing across devices and distribution platforms.”

What this really means is that Google wants to make the web faster, and probably doesn’t trust the majority of sites to adequately speed up their pages or do so on a reasonable timeframe. Thus, AMP were created to allow for pages to load extremely fast (by cutting out the fat from your original source code) and provide an awesome user experience. Users can follow some basic instructions, use WordPress or other plugins, and in practically no time have mobile variants of their web content that loads super fast.

Why use AMP?

While AMP is not yet (or possibly ever going to be) a ranking factor, the fact that it loads fast certainly helps in the eyes of almighty Google and can contribute to higher rankings and clicks.

Let’s take a look at the query “Raekwon McMillan,” the Miami Dolphins second-round pick in the 2017 NFL Draft out of Ohio State University:

Screenshot of mobile SERP for query "Raekwon McMillan"

Notice how of these cards on mobile, two contain a little lightning bolt and the word “AMP?” The prevalence of AMP results in the SERPs is becoming more and more common. It’s reasonable to think that while the majority of people who use Google are not currently familiar with AMP, over time and through experience, they will realize that AMP pages with that little icon load much faster than regular web pages and will gravitate towards AMP pages through a type of subconscious Pavlovian training.

Should I use AMP?

There are rarely any absolutes in this world, and this is no exception. Only you will know, based upon your particular needs at this time. AMP is typically used by news publishers like the New York Times, Washington Post, Fox News, and many others, but it’s important to note that it’s not limited to this type of entity. While there is an AMP news carousel that frequently appears on mobile and is almost exclusively the domain of large publishing sites, AMP results are increasingly appearing in the regular results, like with the Raekwon McMillan example.

I’m a fan of leveraging blog content on AMP to generate as many eyeballs as possible on our pages, but I’m still a bit leery about putting product pages on AMP (though this is now possible). My end goal is to drive traffic and brand familiarity through the blog content and then ultimately drive more sales as people are either retargeted to via paid or come back from other sources, direct, organic or otherwise to actually complete the purchase. If your blog has strong, authoritative content, deploying AMP could potentially be a great way to generate more visibility and clicks for your site.

I must point out, however, that AMP doesn’t come without potential drawbacks. There are strict guidelines around what you can and can’t do with it, such as not having email popups, possible reduction in ad revenue, analytics complications, and requiring maintenance of a new set of pages. If you do decide that the potential gain in organic traffic is worth the tradeoffs, we can get into how to best measure the success of AMP for your site.


Now you have AMP traffic — so what?

If your goal is to drive more organic traffic, you need to be prepared for the questions that will come if that traffic does not yield revenue in Google Analytics. First, we need to keep in mind that GA’s default attribution is via last direct click, but the model can be altered to report different numbers. This means that if you have a visitor who searches something organically, enters via the blog, and doesn’t purchase anything, yet 3 days later comes back via direct and purchases a product, the default conversion reporting in GA would assign no credit to the organic visit, giving all of the conversion credit to the direct visit.

But this is misleading. Would that conversion have happened if not for the first visit from organic search? Probably not.

By going into the Conversions section of GA and clicking on Attribution > Model Comparison Tool, you’ll be able to see a side-by-side comparison of different conversion models, such as:

  • First touch (all credit goes to first point-of-entry to site)
  • Last touch (all credit goes to the point-of-entry of session where conversion took place)
  • Position-based (credit is primarily shared between the first and last points-of-entry, with less credit being shared amongst the intermediary steps)

There are also a few others, but I find them to be less interesting. For more information, read here. You can also click on Multi-Channel Funnels > Assisted Conversions to see the number of conversions by channel which were used along the way to a conversion, but was not the channel of conversion.

AMP tracking complications

Somewhat surprisingly, tracking from AMP is not as easy or as logical as one might expect. To begin with, AMP uses a separate Analytics snippet than your standard GA tracking code, so if you already have GA installed on your site and you decide to roll out AMP, you will need to set up the specific AMP analytics. (For more information on AMP analytics, please read Accelerated Mobile Pages Via Google Tag Manager and Adding Analytics to Your AMP Pages).

In a nutshell, the client ID (which tracks a specific user’s engagement with a site over time in GA) is not shared by default between AMP analytics and the regular tracking code, though there are some hack-y ways to get around this (WARNING: this gets very technically in-depth). I think there are two very important questions when it comes to AMP measurement:

  1. How much revenue are these pages responsible for?
  2. How much engagement are we driving from AMP pages?

In the Google Analytics AMP analytics property, it’s simple to see how many sessions there are and what the bounce and exit rates are. From my own experience, bounce and exit rates are usually pretty high (depending on UX), but the number of sessions increases overall. So, if we’re driving more and more users, how can we track and improve engagement beyond the standard bounce and exit rates? Where do we look?

How to measure real value from AMP in Google Analytics

Acquisition > Referrals

I propose looking into our standard GA property and navigating to our referring sources within Acquisition, where we’ll select the AMP source, highlighted below.

It's only fair to share...Share on FacebookShare on Google+Tweet about this on TwitterShare on LinkedInPin on PinterestShare on RedditShare on StumbleUponDigg thisShare on TumblrPrint this page