In the current state of digital marketing SEO, content is a crucial component for improving a website’s relevance and domain authority within search engines. The challenge, however, is everyone in the know is following suit. This ever-increasing competition makes finding new strategies for higher search rankings a continuous challenge.
AMP has been slow to catch on outside of the world of publisher websites, but as new plug-ins have begun to appear within content management systems, Today’s Business was intent on experimenting with AMP’s capabilities. Accelerated pages looked to be getting preferred treatment in search results, and the opportunity to provide a client with increased traffic was worth the time to experiment.
Today’s Business chose to implement AMP on the blog posts of a local ophthalmology business, with over two years of monthly posts. Pages would be left untouched, since they would require a full redesign. While location pages were optimized and 404 errors were fixed, nothing was done to blog posts on the website besides the creation of AMP versions. There was also no evidence of drastic change in domain authority between the test period and prior. (note: the website was moved to https at the very beginning of the prior control period)
The AMP experiment was initialized with a soft-implementation at the end of February 2017 and then paused after two days. Once satisfied with the configuration, AMP was officially activated on the ophthalmology website on March 6th. The first thing that was monitored was the speed of post indexing by Google.
• 1 Week: 30% of posts indexed
• 2 Weeks: 50% of posts indexed
• 3 Weeks: 57.5% of posts indexed
• One Month: 90% of posts indexed
Based on the stats above, Google is not indexing AMP-ed posts for small businesses as rapidly as it would a publisher’s website. This is not unexpected, but notable when planning a strategy. Next, Today’s Business turned to analyzing and quantifying the impact AMP may have on blog posts.
For analysis, a year-old post about indicators for needing a new eyeglass prescription was chosen for comparison. Since its publication, the post had routinely been one of the most popular ever published on the website, and consequently would serve as a good barometer. After one and a half months, the performance of the blog post was as followed.
During the Test Period:
• Approximately 45% of all sessions for the post were the /amp/ version
o 27% of sessions came from desktop
o 19% of sessions came from traditional mobile formatting
o 9% of sessions came from tablets
• Compared to the previous initial period, the /amp/ version of the post:
o Saw 100% increase of sessions over previous time-period
o New Users increased 65.45%
o Bounce Rate, Pages/session, and Avg. Session Duration all had statistically insignificant declines
What was also interesting was only a small minority of sessions came from the local area. Instead, 56 regions of the world had sessions, including (in order of most to least):
• New York Metropolitan
Today’s Business also compared the performance of all posts to the equivalent previous period for mobile. Branded search queries and posts that were published after the beginning of the test period were excluded.
• 88% increase in post Sessions overall
• 67.75% more sessions from the /amp/ version of the posts alone
Since the initial comparison, a full two-month period has passed since the majority of posts were indexed as accelerated mobile pages. One last comparison was done to compare the performance of /amp/ blogs vs. traditional blogs, as part of the website as a whole. All traditional versions of the posts were included (mobile, desktop, and tablet).
/Amp/ blogs, April 6th to June 6th
• 4.03% of total impressions
• 12.09% of total clicks
• 6.4% CTR 199.5% higher CTR than average page on site
• Average Position 11, 44.52% better than average page on site
• 6.17% of total site sessions
• Bounce Rate 80.17% worse than average page on site
• Pages/session 38.79% worse than average page on site
Non /Amp/ blogs, April 6th to June 6th
• 10.83% of total impressions
• 17.75% of total clicks
• 3.5% CTR, 63.98% higher CTR than average page on site
• Average position 44, 128.7% worse than average page on site
• 8.78% of total site sessions
• Bounce Rate 68.07% worse than average page on site
• Pages/session 28.86% worse than average page on site
In keeping as many variables off the table as possible, Today’s Business has been able to glean some pretty interesting data. The initial results since implementing AMP show notable increases in clicks, impressions, and particularly average position. Blogs don’t have quite the same local optimization found on pages of a small business, which explains some of the difference found in traditional blog performance. However, the AMP blog pages are not only ranking better than their traditional counterparts, but better than the average performance of the website.
The remaining crux is engagement. Bounce rate and pages per session of blogs are typically worse than service pages on non-publisher websites, but because of how much less you can traditionally do within AMP formatting, it restricts the ability to create more engaging style elements. Before assuming basic formatting will suffice, be prepared to spend time strategizing additional amp-friendly elements to bring consumers further down the conversion funnel.
How Google decides to incorporate AMP in the future remains to be seen, but for now, the data indicates that AMP will improve visibility and those who can create quality actionable elements within AMP-ed posts will reap the benefits.