Reading Google’s tealeaves
Still reading Google’s tea leaves?
Try Reading What’s on the Box Instead
Every time Google sneezes, most SEOs catch a cold. Actually, they break out keyword research spreadsheets and retrofit recycled rumours, second hand hints and unconfirmed sightings into their plans. A lot of busy work for all, pouring over tea leaves.
Did January 2016 bring a big Penguin update, a major core algorithm update? Both? Neither? Will April? Are Panda updates being incorporated into Google’s core algorithm or continuing to run alongside it? Google’s not saying… Sometimes Google doesn’t even confirm updates after (we think) they’ve been made. In 2015, of six significant updates, G confirmed three (Mobile-Friendly Update, Phantom 2 Quality Update, Panda 4.2) and the other three were channeled by SEOs.
At the risk of a lot of hate-mail, I say… “It. Doesn’t. Matter.” If you’ve followed Google for more than a few years, its direction of travel is abundantly clear and doesn’t require an announcement about every micro-tweak on the journey. If you’re making somersaults every time G says (or doesn’t) or hints at something, you should pay attention to Google’s trajectory. Google’s announcements should validate what you’re already doing, not what you should begin doing. If your only game plan is to play catch-up with the world’s smartest algorithms, you’ll lose. History is the best (and in G’s case, the only) guide to the future. So get inside where G has been; that’ll be a lot more useful than poring over the dregs of its last cup of tea. Here, in as short a space as I can manage, is Google’s long term trajectory as far as Web sites are concerned:
Google is moving away from pattern matching (matching a query to keywords on a page) towards understanding and anticipating a user’s intent when they make the query and matching that intent to the contextual meaning of a page. As Eric Schmidt and others at Google have said for years, Google wants to provide answers, not just results, and to give you what you want before you know you want it. Google is in the answers business, not the rankings business: Rankings are a means to an end, not the end itself. This means Google getting into the mind of the user (scary) and getting inside the meaning of the page and the site it’s on (easy: right code + right content + right offsite signals).
Directly following on from the above, rankings are becoming a movable feast. Google will put the same page in different results slots according to what it knows about the user, the user’s device and location, the data centre processing the query, and other factors.
Quality. The better you write and spell, the better your legibility and syntax, the better for your search rankings.
Uniqueness. Dump anything that’s cut-and-paste from a content mill, pulled in from a content syndicator, “found” online or grabbed from a stock photo or video library.
Topicality. It’s not just the frequency and freshness of publication, it’s also the relevance to current or trending events.
Internal links must have a reason to exist (contextual relevance) and a logical structure (organic usefulness, if you like).
Link sculpting to boost Page Rank will be less effective over time.
Translation of the above points: Google will treat on-site link manipulation as harshly as it treats backlink manipulation.
You could write a tome on this (many do) but, in summary: Google has always downgraded or penalized links from low quality, dubious or paid sources. Now it will look even more closely at the semantic relationship between the site and page giving the link, and the page receiving it.
Google will likely soon simply ignore billions of links: RankBrain, its artificial intelligence system, which is almost as smart as its world Go champion AlphaGo, now reverse-engineers the relationships between links (giver and receiver) and no longer relies exclusively on traditional link data. If you understand what G’s Deep Mind’s AlphaGo can do, busting bad links is a breeze.
Astonishingly, some people still debate how much notice Google takes of social signals. It’s redundant to even discuss: It’s obvious that Google considers social signals — if it didn’t, then it wouldn’t waste prime search results positions on social profiles — as you’ll notice, they appear regularly in its top search results. QED. The only question is, What part of what social signals count the most?
As with links, don’t buy or rent likes, followers or any other social signals. Natural engagement activity rules. What people naturally do with your content and say about it online, both on your site and off it, on and through their own profiles as well as on yours, will be increasingly determine your rankings.
Earned versus Paid Traffic
A mention in the editorial of The Financial Times or The New York Times is worth 10,000 times more than an advertisement in the same publications. It’s the same with Google’s organic search rankings versus buying an ad next to them: click-throughs on organic rankings are thousands of percentages higher than on the ads next to them, so top organic rankings are gold dust.
Unsurprisingly, given the competition for its prime real estate, Google punishes sites and profiles that look like they’ve bought attention to improve their organic rankings. (That would be equivalent to offering money to a journalist on the FT or NYT in return for a mention). In fact, Google may kick offenders out of the index altogether.
- Being included in Google’s Top 20 organic rankings is like securing the front page or a leader column of the FT or NYT. You have to earn it: talk with a PR adviser.
- Running an ad alongside Google’s top organic results is easy. You bid for it: talk with an AdWords specialist.
- Placing an ad appearing on the front page or an editorial column of the FT or NYT, is easy (but expensive). You buy it: talk with an ad agency.
If you want to get into G’s top organic listings, avoid anything that looks like you’ve bought, rented, Google-bombed or otherwise manipulated traffic to your site or profiles, or have manipulated activity on them. Google will happily take your money for an ad, but you can’t buy a top search-results spot.
Plenty of SME sites are still built on out-of-the-box templates and, worse, dreaded plugin baggage. Google is taking a much more discriminating view of these — for obvious reasons: without a lot of re-programming, they suck for User-Agents and for the UX, especially on a mobile device.
Good server-side and site-wide technology will become non-negotiable, a sine qua non if you want to contend for the top search spots. Up to now, Google has been strangely forgiving of technical incompetence — much more forgiving than we are with sites we work on — but with the proliferation of mobile browsers and devices, Google has decided its time to pull the plug on sloppy back ends. So figure out how your site looks to Google and other machines and clean it up.
- Don’t over-obsess about on-page KW placement. That’s important, up to a point. Instead, look at the big picture and write decent, original, naturally flowing copy — and do it often
- Don’t fixate on ranking for a particular KW. Instead, try to get all your concepts ranking within the top 50 — today’s #37 can become tomorrow’s #7
- Cull every link and social signal that looks like it fell off the back of a lorry or didn’t grow naturally in the wild. You’ll thank me for this
- Junk templates and plugins, they interfere things Google really cares about, like crawling, fetching, parsing, rendering speed and mobile friendliness: 5 excellent reasons to do so
- With social, do it often. Reciprocate. Engage. Build interaction. Use social to drive on-site activity, and vice-versa
- Remember, G is a syntactic machine. In your copy, concentrate on the user, their intent, what’s in their mind. If you do just one thing after reading this post, this is the one.