I would easily take this bet. Today, tomorrow and twice on Sunday. Now, I wasn’t at MozCon, and I didn’t see Dr. Pete’s preso. Just the Twitter stream. So even knowing how Twitter and games of telephone can distort someone’s thesis, I would STILL take this bet. And the reason why is simple, people always overlook local search and the deeper layers of the web. Now look, I don’t need to tell our regular readership that local search is the majority of local and voice searches (which are the fastest growing search types). But I do also want to point out that local search queries are a significant portion of the traditional web as well.
Into the data!
To walk you through my reasoning, let’s dive into the data. First here is Dr. Pete’s data:
Now obviously this isn’t ALL of Google Page 1 (unless the good Dr. got his hands on some sick scraping resources/tech if so hook a SEO up!). He measures some subset of results. Given that the Mozcast looks at the small subsection of the web, about 10k websites, I’m going to go out on a limb and say that he is looking at something similar here. But like I mentioned back at the beginning, the top of the web isn’t representative of the entire web. Let me show you what I mean and dive into the data I’m using to frame my point of view.
HTTPs, in both pack and organic results, is something we looked at in our 2016 and 2017 (spoilers!) Local SEO Ranking Factors. Now I am still working with our statisticians on the modeling for our 2017 factors, but I do have the raw data handy (and have been pouring over it for a couple of months). The chart below shows the increases in HTTPs in both traditional local results and local organic results.
As you can see we are a long way away from a 66% HTTPs adoption rate among the local web. To see just how far off these adoption numbers are, here is a chart overlaying my data and Dr. Pete’s:
For the sake of argument, let’s assume current adoption rate of 11% among page 1 local organic businesses continues. That means we are six years away from a 66% adoption rate of HTTPs among page 1 local businesses. Now I’m going to throw in some caveats. First of all, this data doesn’t look at industries where HTTPs is a compliance issue like financial institutions, healthcare, etc. Also, the local web isn’t the whole internet. Now that being said, outside of the top sites on the web HTTPs adoption is coming along very slowly, and those non-HTTPs sites make up A LOT of page 1 real estate. But hey, don’t take my word for it, just look at the research from our benevolent $GOOG overlords.
Google Research to the Rescue!
Luckily, kind folks over at the $GOOG are paying attention to this, and writing about it in academic, peer reviewed, fashion. Hopefully the aptly titled article “Measuring HTTPS Adoption on the Web” can help us get to the bottom of this.
Per the article, HTTPs adoption rose the most among the Google Top 100 websites. It increased an astounding 20% from 2016 to 2017. This is almost exactly the same as the increase Dr. Pete is seeing among his subset of page 1 results. However, when looking at long tail HTTPs adoption via the Alexa Million the adoption rate wasn’t nearly so high. From 2016 to 2017 the long tail adoption rate was only 5%. There is a lot of page 1 SERP real-estate that these top million sites take up.
Let’s look at http://www.genealogy.com/ for example. First, it’s not HTTPs, and second, it is ranking for 216,000 keywords per SEMrush:
When I filter down to only keywords on page one, it’s only around 200,000
That’s a lot of page 1 real estate that isn’t HTTPs. Now imagine 1 million similar sites that have ~10% default HTTPs adoption rate. And this is in line with what we see in local search. Those businesses that heavily invest in search are adopting HTTPs at a higher rate than those that are less invested. Given the data around adoption rates, there is no reason to think that HTTPs is that meaningful of a ranking factor since the growth of page 1 HTTPs results is more or less in line with HTTPs adoption rates.
Remember, a bad HTTPs transition can bork your site. So if you are thinking about adopting HTTPs strictly for SEO here is a piece of advice for you:
Let the CTO/CIO suggest it 🙂
3 Response Comments
The peril of any presentation is that data points are always going to end up out of context, so first, some context. Yes, my number is a reference purely to page-1 organic results across a 10,000-keyword tracking set. We currently track page-1 Google HTTPS URLs at about 56% of organic results. My “2/3” statement is a projection for our data set based on a fairly consistent historical trend. It does not include local URLs, news URLs, etc.
Our data set is just one data set, but I have seen equally high organic adoption rates in other/different/larger data sets. In detailed analysis, I try to be careful not to dwell on one, absolute number, though and to focus on the trend. A year ago, this same data set had HTTPS URLs at just over 30%. Whether due to pressure from Google or their own volition, many sites are making the switch, including some of the biggest players on the web. I think the trend demands our attention.
I’m not an alarmist (I think my track record speaks to that). Every time I publish HTTPS numbers, I’m concerned that someone might run off and make risky changes to their site. Transitioning to HTTPS can be time-consuming, costly, and carry risk. However, I think the risks of not switching are also very real. Google has made public statements (and I’ve heard even more private rumors) that the Chrome team will be making a harder push toward HTTPS than the search team has. Over the next year, I strongly believe this will expand beyond “pure” e-commerce and into any website that collects data, including sites with lead-gen forms. That means thousands and thousands of small, local businesses. Imagine waking up and suddenly your lead-gen form has a big red browser warning on it — that’s a loss of conversion and bottom-line dollars. As much as I don’t want anyone to rush in without care and thought, I don’t want them to wake up to that situation and have to switch to HTTPS under duress.
There’s certainly a conventional wisdom that adoption is going to slow as we get into the “long tail” of sites. I think that’s misguided, for one primary reason — Google has shown a willingness to simply consume and re-wrap our content. Image search is a great example. Image search has no HTTPS problem because all images are now served directly from Google. From the organic side, there’s a counterpart to this — AMP and any content driven by Google data feeds. At some point, Google won’t wait for us all to switch. They’ll simply take content from non-secure sites and repackage it under their own roof.
I think local is at perhaps even more risk than organic. Google is already pushing us toward Local Knowledge Panels and increasing the friction necessary to go directly to a website from the local pack. There may come a point where, if a site is not secure, Google simply takes search visitors to a Google-hosted local entity and adds even more steps before you can reach the destination website.
Whether my specific number is 57% or 70% by year’s end isn’t the most important point here by far, IMO. What matters is the trend, the freedom that trend gives Google, and the clear intention to push HTTPS across the entire organization (even outside of search team). I think the risks are very real for both organic and local search.
I’m not seeing the same type of positive ranking ‘bump’ as last year, however, at some point, it just makes sense to move clients over to https.
I would take Pete’s side of the bet. Adoption rate will increase as the word gets out even further about the ranking bump an https website could receive. Furthermore (and this is the main reason I would apply my bet), did you know that in October Chrome will start serving “not secure” to any page on a website that has a form? Including contact forms? You better believe every local business in America will be eager to switch to https when their phone stops ringing and finally a customer tells them, “I still found your phone number but did you know I couldn’t access your contact page?” Game over from there.
Also, it’s ANY page with a form. So if you have a quote form on the homepage, it’s going to get served too. Have a search bar in your header or footer? All pages will get served.
Thus starts the mass adoption.