The Future of Organic Search & How TopClick will Lead Client Branded Web-Properties in the Right Direction
The Current Search Climate
Before looking at individual algorithmic changes it’s wise to look at the current climate of Google’s ranking results across the board to get an idea of turbulence:
From Algoroo we can plainly see that SERP turbulence is gradually rising over the years. In July 2014, it was hovering at around 1.5 whereas now in July 2017 (three years on) it’s much more common to see SERP turbulence of 2.5 or higher. This is a tool we have come to trust and has predicted many larger algorithm updates before they fully surfaced (as Google official information) so what we can say is, Google’s search results are gradually becoming less and less stable
Why? If you think about it too much stability is bad for end users. When sites up their prices or change their UI in a negative way, they should be replaced by better websites in Google’s results. In addition to this, Google’s results are gradually increasing in terms of complexity. Remember the term ‘personalised search’? It hasn’t gone away. Google are gradually weighting and factoring more and more user-behaviour and user-data factors into their algorithm. In a way, this makes ‘ranking results’ somewhat of a myth since everyone sees different ones, depending upon the query and user (how much personal data they vent into cyberspace) the results can be radically different. We still use rankings in as pure a form as is currently possible to make actionably accurate decisions, that being said; showing each individual person the information which they require on the spot is Google’s goal and everyone is different.
With a diverse world-audience comes personalisation and as it turns out, instability is a necessary aspect of that. That makes our job in SEO easier and harder simultaneously. You will be less likely to see very stable results which last for months on end. Though that is true; you will also see yourself appearing in ‘bursts’ for keywords which you would not normally get a look in on. Maybe only for minutes or hours, but it will be sporadic and thus the numbers will add up.
In a way, your levels of traffic and users visiting your site shouldn’t be much different, but since Google is rotating results much more on a situational basis – as Google begins to refine that technology (it is currently semi-refined) the same traffic figures should bring users who are more likely to convert. That’s the theory, anyway!
Using Moz’s Mozcast we can see lots of stormy clouds with lightning as Google begins combining and evaluating more and more new technologies into their search results:
So yes, we can say that Mozcast absolutely confirms what the guys over at DEJAN (who built Algoroo) are seeing.
SERPMetrics SERPFlux also shows quite a lot of erratic behaviour in Google’s results:
What’s Driving All the Crazy Weather?
Basically, as we explained on page 1 after examining the Algoroo data:
- New emerging technologies being integrated into Google’s search engine
- Regular on-going updates and upgrades for known algorithms like Panda and Penguin which, whilst they are now updated live with smaller changes (Google Panda and Penguin are part of Google Core instead of being separated out) – still see updates and roll-outs which the community at large does notice as distinct
- Newly created algorithmic aspects which are not yet part of Google Core (like Hummingbird and Rank Brain) – they’re being perfected and amended constantly
- Google’s continued focus on personalised search and improving and refining that technology
- All the usual stuff such as changes competitors are making, jumping ahead of you
Constantly technology is evolving. Recently VR (Google VR) and augmented reality (Google Glass / Google Cardboard) have become pretty big, especially in emerging sectors and industry verticals likeprop-tech (Guardian link there for you) which aims to finalise the technological revolution of the property sector (property is one of the few remaining sectors to be fully technologically realised). Add to this the revolutions of cloud computing over the last decade and how cloud computing might link in with AI (Artificial Intelligence) to create powerful monsters (Cloud AI anyone!?) and you can see that for all areas of tech, these times are filled with flux, turbulence and uncertainty. This is just a fact of the times in which we live.
- Google are invested in VR
- Google are invested in cloud computing
- Google are invested in AI / Rank Brain
- Google are invested in cloud AI / machine learning
- Google are invested in voice activated technology
- Google are invested in mechanical autonomy
As the lines between websites, devices, the internet and real-life blur considerably, obviously SEO adapts. SEO changes over time – as people search differently SEO operates differently.
The issue we have here is that although new technologies and algorithms are emerging all the time, you want to know how we could factually leverage these technologies and algorithmic changes for your benefit. Since these are emerging algorithms and bleeding edge technologies, all we can offer you are our theories. Only older aspects of Google’s algorithm which are tried and tested can yield more input-output type insights from us.
As you well know – anything new involves research, testing and critically risk of investment.
Theories for Augmented Adaptive Search in 2017 / 2018
Here are some predictions:
1) VR and Augmented Reality are On the Horizon
They could still be some years off in terms of wide consumer / searcher adoption however. If you look at Google’s augmented reality apps for phones (or their latest in development iterations) you’ll see that Google are probably moving away from purpose built VR / AR devices and towards releasing software for mobiles to transform mobile devices into augmented reality handsets. Remember the short livedPokémon Go boom?
What a disaster story right? It was a great, fun game and had record breaking adoption upon release. Unfortunately the developers got too holier than thou, blocking cheaters and rooted / jailbroken device users from using their app. Add to this all the bugs, glitches, the disaster of a festival (The Guardian: posted 24th July 2017) and it’s not hard to see why the number of players slumped and it ended up being a garbage heap of a game.
We forget though: upon release before they ruined it – it was magic. People were walking around catching a Pikachu at their local Co-Op, right? Golbat at Costco? Forget about it: it was fun. Now imagine that augmented reality rolled out to general end users, not as a game
Imagine holding your phone up in the high-street and instead of seeing a Pokémon, you see reviews and sentiment scores plastered all over the shops. Gonna’ guess which restaurant to go into in Soho after the Theatre with your friends? Why? Just scan and see what’s good. See which hotels or theatres have tickets or rooms available before you even walk in. This is what’s coming, street-search.
How will we Adapt in Search?
If we need to get out into the streets and plaster walls with QR codes (or whatever the standard becomes) as physical, real life link-based hard-points to your digital properties – then we absolutely WILL. Think a link from the Huffington post is good? What about physical locations which may begin to connect with AR handsets, apps and devices? Most digital analysts and employees consider themselves above such grunt work. We have the foresight to see how physical work could also become digital work as the real and digital worlds merge
We’ll work with you to gain better reviews online. We’ll sit down and work out why some customers are unhappy, change their minds and make them happier. Doesn’t seem like traditional SEO to you? Anything that gains you traffic from Google in any way is SEO
We’ll get some of these devices as consumer adoption picks up. While that square of cyberspace is still freedom-oriented (e.g: any user can change things before it gets abused and locked down with verification) we will BE THERE on your behalf, making sure your best foot is firmly forwards in the VR / AR world-space
How about this for an idea? As augmented reality takes off a real to digital world link standard will emerge (the QR code has a good head-start at this juncture). What if we launch competitions where people have to ‘catch’ physical QR codes in the real world (snap them on their phone camera) to get online discounts? We could put a QR code on a flyer, attach it to a drone and fly it around. That would be a funky PR stunt to pull, as it becomes viable
2) Further Emphasis by Google on Voice Activated Technology
It used to be a social no-no but more and more people are socially comfortable speaking directly to their devices in public. People talk differently to how they type (it’s a less formal mode of communication).
As such more conversational search queries and even direct, actual questions will become keyphrases for which we wish to optimise (much more aggressively). This is somewhat true now but is likely to increase dramatically over the coming years, some people even think that this will make old-school typed keywords irrelevant, effectively resetting optimisation across the web.
How will we Adapt in Search?
We’re ready, we have a number of tools which can easily access Google’s autocomplete data to generate new, more conversational keyword ideas. The first tool to do this properly was Ubersuggest which is quite well known but we use a version of the same technology with wildcard support which enables us to find tens of thousands more keyword variants than our competitors:
We’re ready well ahead of time for the colloquial keyword search-boom which at this point is more a matter of when than if.
Can most other agencies say the same? Do they even know how to search for colloquial search terms using autocomplete data on other databases than Google’s native one with wildcard support? No.
We can get autocomplete conversational keywords and questions from Google, YouTube and Amazon already: re-firing the non-Google keywords through the Keyword Planner to see if they have Google search volumes. We’re ready and prepared.
3) Google to Refine Cloud AI / Machine Learning
Rank Brain is not the end of the story – if anything it’s a relatively crude and simplistic overture to what is yet to come. Machine learning and cloud AI will be used to spot bad, unsatisfactory results before they are even served to users. What Penguin and Panda started (a focus on value-add marketing efforts over manipulation) AI will surely finish
AI is exactly what it says on the tin. Instead of relying on ‘clever’ algorithms which are essentially very detailed processes, Google will be able to ‘think’ more like you and I do. With quantum computing making leaps and bounds, it’s only a matter of time before AI takes centre-stage. If cloud quantum computing becomes a big thing, that will be that as they say.
But doesn’t this open up better avenues of ethical marketing for us?
Before if Google’s algorithm arbitrarily decided that we had done something wrong, we had to wait hours (or even weeks or months) for a manual spam action employee to review our full case study. In the future that won’t be a problem – penalties will still happen but we’ll have an artificial intelligence to debate with and persuade. We haven’t had that before so actually that’s great!
How will we Adapt in Search?
Currently Rank Brain pretty much exists to serve search results to users who are processing search queries which have never been asked before. Annually this accounts for roughly 30% of Google’s searches (yes even still!)
One way in which we could leverage Rank Brain now is to predict keywords which will have volume in the future, which as of now have never been asked. We could then optimise for those queries, safe in the knowledge that initially Rank Brain (and eventually Google Core) would weight our content heavily. The problem is we have NO WAY TO KNOW which queries have never been asked, as Google’s keyword planner gives zero for any search term(s) with less than 5 average monthly searches (it rounds up and down in 10s). We might be able to use Bing data (which is actual and factual) to have a good guess.
The problem here is that without a crystal ball, we can’t know which search terms will really take off so it would be very hit and miss.
Although we have an option here – we would not advise using it.
Instead: let’s wait for Google to deploy AI to other more interesting areas of their ranking and penalty based algorithms. For this technology – it’s just a waiting game
4) Google to Enhanced Organic Retargeting / Personalised Search
As we know Google is always looking for ways to use personalised search to enhance their search results. Personalised search can be roughly defined as the algorithms which move a user’s search results away from Google’s pure, default ranking algorithm – tailoring them to the specific end user.
Personalised search can alter many aspects of a user’s search results:
- The order of Google’s organic search listings
- Whether users do or do not see various ‘universal search‘ injections (news results, image results, video results etc)
- The order of universal search elements if present
- The context (query type) of a given ambiguous search query (e.g: “protein shakes” – is the user a regular buyer or do they have no past history of purchasing health and fitness products at all? Is the query transactional or informational? Do they want knowledge or products?)
- The appearance and placement of knowledge graph entries (where Google renders the information / answer to a query directly in its search results, above or to the side of the usual list of search queries)
- The appearance and ordering of site-links appended to individual Google organic results. For example, if you search for a brand which supplies multiple services, you get generic site links but if you also specify a service sometimes you get the page for that service pushed higher up the site links (or you get just that URL and no site links). Personalisation can also affect what site links people see
- The SERP snippets / rich snippets which a user sees. Usually if well written or correctly implemented, this will be your Meta description with possible schema / microformats / microdata information (reviews) appended
- The appearance or ordering of related search results. Since a search query can be ambiguous depending on the user, the context of the query and therefore the stream of related results can also flux.
- Many other (rotating or in-dev / testing) factors!
Factors which are inputs for personalised search may include (for any given query):
- The user’s IP address
- The user’s geo-location
- The user’s past browsing history, cookies and / or sessions
- The user’s Google search history which is cloud (not browser) based and separate
- The user’s web browser and / or user-agent
- The user‘s device (PC vs Mac vs Samsung Galaxy S7 vs iPhone 6 etc.)
- Whether the user uses AdBlocking by default or other niche tech such as forcing HTTPS on all sites. Unique browser setups, plugins or add-ons for the user
- Many more factors such as the user’s YouTube search history and appended data
How will we Adapt in Search?
To a large degree we have already adapted by increasing our knowledge of structured data and the ways in which personalised search data can impact how our search results appear to end users.
Structured data gives us slightly more control (very situationally under very specific circumstances – it doesn’t always get used contrary to popular belief even if implementation is perfect) over how search results appear to end users.
Contrary to popular belief schema isn’t the only form of structured data processed or understood by Google, we also have familiarity with and are aware of:
- Microformats (sort of an umbrella term)
- Microdata (a designated specification)
- Classic contextual / semantic HTML tags (e.g: <H1> is pre-HTML5 yet is still semantic, it designates a header. Stuff from HTML4 and earlier)
- HTML 5 semantic tags (same as directly above but newer)
- Classic structured data (e.g: <ul> / <ol> / <li> to designate lists and <table> / <tr> / <th> / <td> to define structured tables of data)
- Structured data feeds (e.g: XML, RSS and atom formats and standards)
- The structured / nested nature of HTML itself
- Arrayed data (such as structured data presented in a JSON array – ONLY if Google decide to enable rendered crawling for an individual web page in the specified crawl instance – they don’t do this very often as rendered crawling takes like 10x longer than DOM / source scraping)
- Niche / specialist structured data (such as WAI ARIA which can make browsing the web significantly easier for blind or visually impaired people)
We also know that results are hugely impacted by the external, un-biased popularity of listed web pages and resources. As such backlinks and social amplification have a huge impact on the way in which Google adapts its search results for individual users who are flagged as part of a demographic with certain interests and hobbies (you can actually find this data for your audience in Google Analytics).
Weaving together technical and inbound marketing practices and factors, we will of course try to leverage personalised search at every opportunity to ensure that those who are often or always served your search results are relevant users who might be up for a spin or two.
We will engage in more audience research and idea-based outreach campaigns than ever before to get the right search users to our client’s sites, to make personalised search a boon rather than a burden.
5) Google’s General Goals & Our Continued Universal Adaptation
Time and again Google have stated that they focus on their users and providing a positive experience for their searchers:
As such instead of performing old-school SEO tactics like directory submissions and PageRank sculpting (which don’t even work anymore anyway) we focus on VALUE-ADD MARKETING
This means that we (like Google) focus on the value-add of our changes and activities with regards to the user who will be digesting our ideas, promotions, content and technical changes.
This necessitates that we have a heightened sense of what the world ‘relevant’ means. For example: placing a link within an existing post on an automotive magazine to a car insurance product elsewhere on the web is linguistically or thematically relevant (both sites are about cars) but may not be semantically deep-relevant. Is it relevant for the user to click on the link? That’s the real question.
If the link simply points to a default landing page where the car insurance brand dictates their price (which may not be the best price) to the end user, it’s not relevant as a link to Go Compare or Confused.com would have far more ‘value add’ for the end user. If on the other hand; the link points to a page which knows the active brand’s lowest price on comparison sites and then additionally applies and extra 10% discount as part of the promotion, it could be relevant (though ideally in this scenario the article would also have to be fresh not historic).
If a discount were made clear it would obviously attract many users to click through the link, transforming the link into an editorial traffic carrying link rather than an advertorial ‘for the sake of SEO’ link. Something with value add as part of an editorial piece is acceptable.
All of our activities have to add something new to the web in the form of digestible, digital value for end users. The best definition of value-add in terms of content came from the BBC’s founder John Reith which is that we should seek to:
As long as our content is doing at least one of these things well (hopefully two or more though) and as long as any links back to the mothership are citations rather than advertorial links (which would have to be no-followed) then our content and link(s) will pass as editorial. That’s what we have to do now.
Many SEOs mistakenly roll around the web spouting that Google has banned or doesn’t like guest posting, this is factually untrue even though you can find examples of Google stating such – as those comments from Google are simply Google reflecting the SEO industry‘s own misuse of language back at us.
Because a lot of ‘SEOs’ are very lazy what began as a legitimate practice ended up with SEOs paying the owner of a web-network to pay their subordinates to write thin content and place it on random blogs across the web. These are not actually guest posts (which are editorial in nature) they are sponsored posts (which are advertorial in nature). It is the presentation of advertorial content as editorial content which Google highly frowns upon.
As long as we treat our content as editorial (produced factual or opinion pieces, or games or interactive HTML5 / CSS3 infographics with something real to say) then we can still perform outreach, ideas generation and guest posting. Sometimes (often) a fee will be charged to place editorial content so that the editor can editorially review the content and see if it is fit / suited for their audience.
Here is the Basic Format for How Links Work in 2017 / 2018
- Write a detailed, data-referencing study which says something interesting and host it on a client-branded blog or news section
- Transform the detailed article into a more emotive opinion piece with a contained infographic, fully illustrating the data in an entertaining way from the main article
- Offer the article and appended infographic to webmasters or editors
- Pay the editorial review fee
- Get the content listed externally, linking back to the main detailed article / data-piece as a citation rather than an ad-based link so users can see the detail and data to verify what they have just read or seen on the external site
- Soak up any links which may happen to spawn across the web as a result of the great content backed by a factual study or a promotion which would factually benefit the user to an extent which they could not be benefitted elsewhere online
- Hopefully the users on the external site are informed, educated or entertained so our link remains perceived as an editorial link by Google
- Ensure the link is not no-followed which it shouldn’t have to be, as it’s editorial
- Ensure we don’t end up with a big ‘sponsored‘ label across the content
We no longer do ‘clever little things’ for £5 on Fivver which boost rankings through ‘voodoo magic’. That epoch of SEO is dead and gone
SEO (like other historic marketing verticals such as PR / the printing press) is now highly competitive. With this maturity comes the unfortunate fact that more factual effort must be invested on both sides(ours and yours) in terms of analysing trends and generating content. Implementing technical changes (as inbound metrics are now infinitely less easy to game) is more important than ever, this means that when we recommend technical SEO changes, in this new age they really just need to be carried out without hours of back-peddling and debate.
If we continue to look for ‘magic bullets’ and ‘quick wins’ with no effort invested, we will assuredly fail. If we embrace value-add workflow and try hard together- we can only succeed!