A couple of days ago, I was chatting with a buddy, talking about the difference between ranking signals and ranking factors. It got me started on a train of thought, and as I jotted down a list of a few factors and signals, I started thinking about what the criteria might be that define something as one or the other.
I had a fair good idea of what each variable is, but I found it difficult to nail down a clear definition that would readily classify something. So I went to Google and started doing some digging.
Ranking Factor or Ranking Signal
First, I looked for obvious things like ranking factor versus ranking signal. That came up dry. Seems like nobody in the first four pages had ever clearly defined them, let alone Google (big surprise there, right?)
So then, I did some more extensive searching, for terms like:
ranking signal
search ranking signal
search engine ranking signal
search algorithm ranking signal
followed by repeating the same searches, but substituting factor for signal. Then I looked for common results. By that time, I wasn’t too surprised to see that with the exception of a few documents trading positions, the results were pretty much the same.
So I chatted with a couple more friends, getting their opinions, and one of them told me that he had once asked a Google engineer at a conference what the difference was between a factor and a signal. The answer they gave him was that they’re the same thing.
That doesn’t ring true to me. I’ve always looked at ranking factors as a fixed attribute, such as on-page elements or number of links. Signals, I’ve always thought of as being more liquid, such as social media interaction or freshness.
That’s when I decided this could be an interesting poll. Not that the results will yield any definitive value… I’d just be interested in seeing what other search professionals think about it. So I built a short list (not all-inclusive, by any means) and built a poll, so that you can simply check the things you see as ranking factors. If you think something is only a signal, leave it blank.
The poll is simply one question: Which of the below are direct ranking factors (as opposed to being only signals or having no effect whatsoever on ranking)?
I’d appreciate your input. The more people that respond, the better. It should only take you one to two minutes. I’ll be publishing the results next week.
Take our Ranking Factors Poll
“That doesn’t ring true to me. I’ve always looked at ranking factors as a fixed attribute, such as on-page elements or number of links.”
Your preconceived notion notwithstanding, it’s not up to you or me to decide for someone else what they mean. A now-retired academic researcher, Donald Swanson, noted in 2000/2001 that “undiscovered public knowledge” arises from the wide dispersal of research resources (Cf. http://www.asis.org/Bulletin/Mar-01/swanson.html), leading to sometimes complementary but distinct points of view.
We would call those distinct points of view “preconceived notions” but in reality we are each defining our own languages to discuss the same topics, and so when we hear something from the other side it sounds strange. It’s better to learn to think in terms the other guys are accustomed to using rather than to suggest they are being disingenuous.
No preconceived notion, Michael… simply my interpretation of things that Google says, seasoned with my own observations. For the record, my “doesn’t ring true to me” remark wasn’t meant to imply I thought the Googler’s response was lacking in veracity… simply that my impressions lead me to believe there is, in fact, a subtle difference.
A friend put it much better than I did:
“The only meaningful distinction is between those whose weighted effect is and is not independent of other input variables.”
I think the “independent of other input variables” sums it up very nicely… much better than my “fixed” vs. “liquid”.
I think they use everything in some capacity as long as they own the hardware, software or spyware 😉
LOL. Probably closer enough to the truth to make some folks nervous, Deb.
Interesting choices of voting options Doc, sifting the signals from the correlations reminde me of walking along the edge of a cliff blindfolded.
I’m not really trying to sift signals from correlations, Mel. My response to David may clarify my thoughts better.
Sorry, but I’m with the Googler on this one. If Google uses it in its algorithm, it is a factor. Or a signal. Or a piece-of-the-pie. Doesn’t matter what label you give it, it is either in or out. There is no middle ground here.
Thanks for chiming in, David. That used to be my opinion, too. At least until I started to think more along the lines of layers. But I’ve come to think that some things are looked at more in how they affect a site’s ranking when combined with other things, rather than as stand-alone attributes.
For instance, the effect of most on-page issues, such as keywords (or recognized substitutes) in the title, can be more or less quantified by an algorithm. I could see that being given a certain “value” on its own merit.
But for other things, such as the presence of a KW in the content and its proximity to other terms that might modify its importance, quantifying their effect may be more inter-dependent with other “layers”.
Using an analogy of mixing paint pigments: yellow and blue will yield green, and adding red will give us brown, while yellow and red alone will produce orange. None of their effects can be readily quantified by themselves. That’s somewhat the way I envision signals.
Granted, I may be out in left field here, reading too much into statements by various Googlers. But pointed questions on a “factor” often bring measured responses changing the terminology to “signal”… so often that it’s difficult to ignore. That’s what creates the impression that there’s a difference, at least perhaps, in independence.
A fine point, I suppose, but given the tendency of many people to focus on a single “shiny thing of the month”, it may be a distinction worth thinking about.