SEO News

Posted by randfish

For nearly two years, marketers have been frustrated by a steadily increasing percentage of keywords (not provided). Recent changes by Google have sent those numbers soaring. The site Not Provided Count now reports an average of nearly 74% of keywords not provided, and speculation abounds that it won’t be long before 100% of keywords are masked. Without that referral data, our tasks as Internet marketers become far more difficult—but not impossible.

In this special Whiteboard Tuesday, Rand covers what marketers can do to make up for this drastic change, finding data from other sources to stay on top of their SEO efforts.

For reference, here’s a still image of today’s whiteboard!

Video Transcription

Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don’t have time to hunt down but want to read!

To illustrate the point, here’s our data from August, showing (not provided) as a percentage of all Google organic traffic:

And the same chart for September, so far: 

And for today: 

Other sites are reporting similar patterns. The Mirror’s Malcolm Coles reports that of The Mirror’s desktop traffic from search, 82.5% is encrypted today.

Koozai’s Mike Essex reports that his site’s (not provided) traffic is 93% of organic, while the (not provided) count, which tracks a number of sites, shows a big spike over the past couple of weeks, though this doesn’t include today’s data: 

Why does this matter? 

For us, we’ve kind of given up on making sense of organic keywords, simply because we can see so few of them. 

Here are our top organic Google searches for September. As you can see, the numbers are fairly insignificant. All it tells us is that we’re doing well for ‘Bill Gates quotes’ thanks to a seven year old article. 

Even this workaround for calculating branded search traffic will soon become very difficult, as we have little or no organic data to base it on. 

Of course, search data isn’t encrypted for advertisers using Google search ads, so if you’re worried about the NSA spying on you, don’t go clicking on those ads… 

Check your (not provided) traffic

These handy custom Google Analytics links from Dan Barker will help you to quickly check the state of your (not provided) traffic. 

• This dashboard shows Not Provided as a percentage of Google Organic traffic in pie chart form, as used above. Click here to add to your GA profile.
• This custom report shows the same, with keyword data, as in the screenshot above. Click here to add to your profile. 

Are you seeing a jump in encrypted organic search data on your site? Let us know in the comments… 

Personally, I think most commentators are missing the bigger picture. I am by no means right all the time (or even a small amount of the time, to be honest) but every now and then I send out a spark that ignites something that goes on to become a raging inferno. 

Over the last month my brain has been largely filled with thoughts around authorship and how Google could use it in the future.

My main argument is simply that authorship needs to be seen as a potentially fundamental change in the way that Google assigns value to content on the web.

Here are a couple of caveats before I get started:

• Firstly, I don’t work for Google.
• Secondly, I’m not forcing anyone to do anything or implying that this is happening yet. I’m describing a potential trajectory that digital marketers need to consider.

Authorship as a value metric for links and mentions

As SEO evolved over the last ten years, different metrics have been used to identify the value of an individual link. Originally we used PR (PageRank) as this was known to be a metric Google used to measure the ‘power’ of a site.

We then moved away from PR to a more authority-based approach, taking into account the perceived importance of a website within its niche and combining it with MozTrust and DA.

This means we have moved from ‘pure power’ to ‘authority’ already.

Now we are seeing Google pushing authorship as a way of getting content noticed (it allows you to have a picture next to your listings in SERPs), but as with most things Google, there is always some underlying motive beyond the direct benefits to authors visibility.

Firstly, authorship is yet another attempt to push content producers on to Google+. Google+ isn’t going away. That’s a fairly obvious motive for Google’s authorship push, but could there be more? 

Could authorship be used to further segment a website into pages of greater value and pages of lesser? 

At present we focus on the authority/trust of a domain in the knowledge that trust is applied at a domain level rather than a page level.  

This assumption of domain level trust is perfectly valid in that the seed sites for a trust algorithm are so well researched they are assumed to have no areas of lower trust.

But is applying trust at a domain level the best way to go? Well, not really! Recently, and not so recently, we have seen a lot (a real lot) of activity around newspapers and trusted publishers trying to make money online through selling links: 

• Many come from tried-and-tested advertorials.
• There is no doubt that money is changing hands in exchange for mentions and links directly from journalists, but this is very under the radar;
• Services such as HARO are now widely used by PR and SEO teams in an attempt to gain links.
• I recently overheard a group of students talking about how they were all doing travel internships with the aim of becoming travel writers because they can get lots of free holidays, etc. This activity of ‘I give you a holiday, you give me a link’ happens very often. 

If you were trying to prevent the above from having such an effect on the overall trust allocation across the web, what could you do? Well, it’s easy: you take some of the power away from websites and give it directly to the authors.

This is because: 

• Advertorials don’t come from authors, as such.
• High-value authors are less likely to take back-handers, or are at least so expensive that the activity is very limited.
• The best authors don’t use HARO – they do all of the work themselves. The middle and lower tiers use these services to supplement their workload.
• The best authors and publishing organisations send their own staff to places; they don’t rely on touting for freebies.

If Google was able to identify these ‘best’ authors somehow, it could use the data to apply an additional level of trust to the ranking algorithm. They could also use the data to police author-commercial relationships.

Hypothetical scenario number one

Bob is a well-known author and writes for his own personal blog and a well-known tech publisher. At present, the tech publisher would be the target for a link, as it has the higher DA.

Bob’s own blog, although awesome in terms of content, doesn’t have a very high DA.

If Google were to be able to flag Bob as a trusted author in the field of tech, suddenly anywhere that Bob writes about tech has a much greater value.

Hypothetical scenario two

Bob has an intern working with him as a writer for the well-known tech publisher. At present it wouldn’t matter if the intern or Bob wrote the content as long as it was on the site and had a link. Both would result in the same Domain level trust flow.

If Bob has been identified as a trusted author and the intern hasn’t, then the link suddenly has a lot more value if Bob wrote the content compared to the intern (I wish I had given him/her a name earlier). 

Obviously the important stuff comes from blending the two scenarios together. Could we get to a stage where, in terms of value:

• Bob writing on the tech site > Bob writing elsewhere > Intern writing on the tech site > General link from tech site.
• Bob has the power (you won’t hear that said often).

Would this be open to being gamed?

The SEO community will have a damn good go at gaming anything it can, but when you really think about it the only way to game authorship is to basically buy trusted authors.

This happens, of course, but the sums needed to do it successfully are putting it out of the reach of most people’s marketing budgets.

You don’t just become an author by writing a 500-word article on a random blog and getting 500 paid Google+’s out of it, you have to work bloody hard at it, and also maintain your status over time.

There is no obvious way to automate becoming an author other than maybe hacking into trusted authors’ Google+ accounts. 

If this were true, how would it impact the way SEO works? 

If you are doing things properly (subjective, I know), it wouldn’t. You should be generating good-quality content that talks to your consumers and then seeding it out to influencers.

These influencers should naturally include trusted authors. If anything, it would make success easier to report. At present, ‘Bob wrote about it on his personal blog’ is a much harder sell than ‘Bob wrote about it on TechCrunch’.

Authorship as a counter to guest blog posting

‘Directory link building is dead!’, ‘reciprocal link exchanges don’t work!’,  ‘paid link building is high risk!’, ‘infographics are (insert expletive here)!’…  As SEOs, we have a tendency to find something that works (in this case, something gets links), and burn it out through mass usage.

Having just returned from Brighton SEO, the in topic at the moment seems to be guest blog posting, or content provision. 

Now I’m not against guest blog posting, though I had some thoughts about its future, taking into account the likelihood that at some point Google will look to devalue or make poorly-implemented guest blog posting toxic.

The general concept of guest blog posting

SEO, link building, outreach and off-page: it all comes down to getting a link from a domain. I have done what is needed to get a link from site A to my client, move on to site B.The way we get the link is largely irrelevant.

We give them some content, we commission them to write some content, we give them a product, we provide them with an infographic, and so on. 

This approach, if carried out so casually, leaves a footprint that’s very easy to spot: brand X is mentioned in one post on Y number of sites over Z period of time. Often, none of the posts have any social metrics associated with them, no comments, the writing doesn’t follow the style of the rest of the site, etc. 

Still, this isn’t a post about guest blog posting, so I’ll take this opportunity to move on.

Enter authorship

Authorship is all about being a contributor of merit. Let’s say that Google sees this as someone who has made a genuine effort to contribute to the site in question beyond simply writing 200 words with a link in it.

The author has a profile. The author has a history of posts for the particular site. The author has value in the particular niche. The author engages and promotes their work.

If we compare this to a 200-word post stuck on a site by a guest author as part of a guest blog posting campaign, we can see a distinct difference in activity and the associated footprint.

Posts from genuine authors and contributors have value. Low-value guest blog posting campaigns don’t.

Suddenly, you have to be genuine, you have to build a relationship, a history and much more. Everything gets harder which, at the end of the day, favours Google and the people willing to put the effort in.

We contribute either through producing something worthy of mention or by becoming a genuine contributor to a site – no more ‘providing 200 words’.

Would this be open to being gamed? 

Hell yeah, but it would be a darn site harder to game than the way guest blog posting is going now. Getting an author to write for you is nothing new, but with authorship they are going to have to take more control of how they are seen online.

We are likely to see ‘toxic authors’ who are known to be operating outside of Google’s ideal of a valued author. The cream will likely rise to the top.

If this was the future, how would this affect the way we work? 

‘Guest blog posting is dead – long live expert author creation and contribution campaigns’.

If authorship was used as mentioned, the equity passed from site A to site B would be based on the contribution of the author overall.

As an example, Kevin Gibbons has made a solid contribution to Econsultancy, posting over a period of years on various subjects all relating to online marketing.

If Kevin links to you from an article he wrote on Econsultancy it is likely to get shared and commented on and thus is likely to have more worth. 

If a random author (Malcolm Slade, for example!) was to link to you, people would probably think ‘who on Earth is this guy'” Rightfully so, as I am not a valued contributor on Econsultancy. It’s how humans work, and that’s what Google is always looking to replicate. 

There is a simple lesson to learn here. Don’t build links: contribute. Find the sites your client should be involved in (dictated by their strategy and by customer insight) and get involved for the long haul.

Think like a PR would. Build real relationships with real authors and help them to help you. 

Please feel free to discuss, comment, share, argue or rant. I appreciate I miss-use quotation marks and this could have been two separate posts but if I have got you thinking, I’ve done what I set out to do.