It is pretty much a meme now for SEOs to share and collectively make fun of bad hot takes about how “SEO is dead” by people who don’t even have a baseline understanding of search.
They (we) think, as long as people look for answers and things, search will be forever important, and as a result, effectively optimizing for search will be a thing that knowledge workers do.
Deep sunk cost bias? idk.
But href links are actually dying. Maybe it’s the way “we’re all dying,” but still, I would eyeball the value of a typical a href link in ranking a page diminishing 3 to 10% decay per year.
We all forget – href stands for hypertext reference. A link is a conveniently clickable reference to something else in html. In a sense, it represents a meaningful relationship between two pages, a relationship that has discernible context and value.
The page says to the user – “hey, you’re here for me, but look here is this other related thing, you can leave me and go there.” If you love the user set them free. If they come back it was meant to be. Swoon.
What a powerful recommender. And so now we have Google.
Link building approaches were pretty much all conceived during a time when links were the strongest signal Google’s search engine relied on to determine quality.
We all snickered when circa 2014 G announced they tested a version of the search engine without links and (weirdly also announced) that it went poorly.
Central to their web spam mitigation strategy, engineers have been working for over a decade to get away from dependence on link signals. Now it’s 6+ years after that announced experiment and I think they’ve got it figured out.
No one worth their salt thinks of ranking factors as clearly defined hard edged things like “links!” anymore.
It is a full color spectrum spanning multiple dimensions of context, to name a few:
- the users’ context in the moment accounting for a previous search in a new search’s results
- the users’ general context personalizing based on past search behavior
- how timeliness influences the context of meaning of a search, think “sandy hook” after Sandy Hook happened
- the context of physical space and geography
- the context of underlying meaning of words in the world given Google’s Knowledge Graph and robust awareness of how things and ideas relate
The multi-dimensional context has become so incredibly… multi-dimensional, I honestly wonder why they need “links” at all.
G’s newish PR vibe is that they “still use links to find new pages.” The subtext – they use links for indexing and that’s about it.
It is hard to believe they are saying what they mean, when they never clearly say what they mean and every other Twitter retweet or reply from their main point person for questions reeks of sarcasm (sorry, John!).
It’s reasonably clear that there is more to it than links for indexing. Link quality is still a strong correlate to rankings. But things can correlate and the reasons they correlate change over time, e.g. better proxies being used.
A longish-term study by Stone Temple, now Perficient, shows stronger correlations for financial industry links than others, like medical. Link value is likely topically context dependent.
There is also SERP diversity to consider, the idea that many people can search something and have an entirely different intent. To return a diverse array of results, it stands to reason backlink profile diversity of results should be factored into creating that array. But that’s again another correlate.
NLP has lots of ways to discern homonymy and polysemy from word context and G has always been on the forefront of that.
And they’re not just turning down link signals by turning up other signals for organic results, it goes well beyond that – think zero click results with knowledge panels, news cards, youtube videos, they are hitting this from all directions.
Anecdotally, I see too many cases where a client (or client’s competitor) knock a piece of content out of the park, with minimal content amplification, and outrank some mega linked to article handily.
On the flipside, I see too many cases where a couple good not great links don’t move the needle on a piece of content at all.
That’s all a round about way of saying: G has been steadily building a world where other proxies for relevancy and value can be used. Whether they fully get there is irrelevant.
The decay of the “a href” is real, and moving that needle is central to Google’s survival strategy as the search engine that can’t be manipulated (so easily).
How many other activities are you still in the habit of focusing marketing energy on… even though the returns are also getting worse and worse each year?