How Does Google Measure Quality
By Mark Ewbie
How on earth can Google tell the difference in quality between one page and another?
That would be a heck of an algorithm that can read my page or yours, and form a judgement based on the content, humor, entertainment value.
When even literary critics can't agree on a review for a book - what possible chance does a dumb, although clever, bit of software have?
Caveat. This page consists of my opinions – as valid or not as virtually anyone else’s on the net.
It is my own view on how Google measures my articles against searchers expectations and experience, and ranks me accordingly.
It may be entirely untrue or misguided, but I’m prepared to put my money where my mouth is, and use this method or guide for my articles.
If you are thinking to yourself “what the hell does this idiot know?” then that is a perfectly reasonable and sensible attitude to take to any amateur, or indeed, professional writer.
Make your own decisions in life.
I am making mine.
My Experience as an Article Writer
I have been writing articles for no money for about a year now.
I have not cracked keywords, search traffic, affiliate sales or any of the necessities to build that mythical passive income.
However, I can’t quite shake off the hope that there is a slim possibility, and for now that is enough to keep me ploughing on.
Slowly and painfully, well slowly anyway, I am finding out what works for me and picking up a huge amount of misleading and damaging advice as I go along.
Unlike the masses of thin articles trying to sell you secrets, or ebooks, or can’t fail opportunities – I am really thinking aloud in this page.
It is my approach which I plan to follow for the next year. Or longer.
Seeking Traffic
It’s a simple goal.
The more traffic you get to your page the more money you make.
Since Google and other search engines started people have been trying to crack this challenge.
It’s all very well researching a keyword and writing an article, but how to compete with the thousands of other similar pieces?
I noticed yesterday that eBay have adverts for people selling Google Likes.
So you can buy two hundred and fifty Likes and push your page up the rankings.
This is the same thinking that gave us years of linking games.
In my opinion, although I have already pointed out this is my opinion at the top of the page – but you know, people will get offended…
Linking and Likes
… Linking and Likes will not reward with traffic. It is so easy to game them that they become meaningless as a measure of anything.
Good! I am so hacked off with the people selling fake links, and because someone bought more than me they get a higher ranking.
Now I know Google doesn’t care whether I’m hacked off, but they do care about the reader, the searcher.
In a world where Links and Likes are freely available – how much credence do you think they attach to them as a reference of quality?
Not much I suspect.
How do I know this?
I don’t know for sure. I do know though that I don’t back link, or do anything at all with Likes, and neither do most of my readers (they don’t Like me) – but I do rank OK for some pages.
Remember what I said about forming your own opinion? It’s important.
I believe the purpose of Likes is not actually to tell Google that the Liked page is really good, but for them to form a view about the Liker. So they can better target marketing and advertising at them.
He Likes pages about that subject therefore we will tweak his web experience (ie. advertising) slightly.
When I say Google Likes, I mean +1 of course, but we all know it’s just another Like.
The User Metric
What would be the single most important indicator that a page matched a reader’s expectation?
Not necessarily quality. Google can’t judge quality, no one can.
One man’s favourite page is another man’s piece of rubbish.
Their aim is to try to give the reader what they are looking for with a minimum of searching.
You may mock, and say they are rubbish at it, and that Panda hates us and we can’t understand what they are doing…
… but that is their aim.
I believe that more than ever before, Google are measuring searchers page activity and using it to tweak their rankings.
Each time a reader stays for a while, maybe uses one of your on page links – that is a plus point. The opposite of that is a quick back arrow and a failed search.
If there are a thousand pages on a particular phrase, Google will randomly let them all have a taste or share of search traffic.
The occasional hit now and again for each page.
Using the responses from those hits it can shuffle the pack of a thousand articles, so that gradually the cream rises. I believe this is happening with greater frequency than a monthly deep crawl.
In a fast moving world, why would Google want a metric that is up to a month out of date?
What Does This Mean?
If you buy my assumptions above then that old chestnut about quality might finally be coming true.
Those pages that satisfy a searcher will do better than those that don’t.
This seems perfectly reasonable, achievable and desirable to me.
Since starting writing I have hated the gaming idea, it seems unfair on the ordinary writer who is putting their heart and soul in to creating something of worth only to be outranked by those who know how to play the game.
Wishful thinking?
That may be so, but it is my thinking and will govern my attitude to article writing.
I am now revisiting everything I have written, with a view to ensuring that the content matches what someone pulled in by keywords and title would expect.
Next time Google allows me a hit on a page I want that searcher, if at all possible, to be satisfied with the contents.
And I’m keeping my fingers crossed.
Am I Just Guessing?
Well yeah. Google didn’t write to me personally and explain how I should approach internet article writing.
However, I do spend a LOT of time looking at every statistic I have available to me.
For each article I check where it sits in the rankings – page one, two, etc.
If I get a visit I look at the bounce rate, time on page, keywords used.
I think about that searcher.
What were they looking for and what did I give them?
And I do the above things every day.
Obviously I can’t do all two hundred articles. Most of them don’t get traffic – at all. I’ll get around to them.
My main concern is the ones that do – because I want them to carry on getting traffic.
Personal Note
I started, and continued, writing as an ego filled humorist.
I like nothing better than creating a misleading spoof with a dash of humor.
Unfortunately that means a large percentage of disappointed searchers.
It’s a very small number who will be delighted to land on some writing that purported to be one thing, and turned out to be something else altogether.
I have a lot of titles and pages to tweak and rewrite so they match more closely with searcher expectations.
Those of you who have written good content, accurately targeted at search visitors will have a deal sight less editing to do.
More Caveats
I’m assuming in the article above that you or I are not subject to a Google or Panda slap.
That is a different area.
In that case Google has made a judgement about your site based on…
… well ‘based on’ is subject to a whole load of rumours and misleading information again.
Possible areas could be dodgy linking strategies, keyword stuffing, thin affiliate content, copied content – or a combination, or none of those things.
My views are written for a non slapped article writer wondering how to get and satisfy readers.
Me, in other words – but you are very welcome to have read it.
Warning: Mark Ewbie is not an SEO expert. He has not written a book. He was never employed by Google or any affiliates. The views in this article are his alone, and may be the ramblings of a confused man. Do not give him any money.
Comments
Mark, I think people impacted by Panda would be well served to read this. While I'd need to write more than a comment to get my complete point across, I believe Panda/Google is about the query the user asked and if they were satisfied. If Google thinks your content doesn't satisfy readers, then bam!
I also think that when a site get's hit by Panda, Google leaves a trickle of traffic that is the clue. Do you get referrers from Google where your site doesn't answer the question? If not, fix it so that it does or take other steps.
This is funny Mark and I think you have some good ideas here as well. Up, funny and useful.
Snakeslane, you are much too nice to slap, plus you own a lucky magnet which can ward off evil spirits.
Paul, thanks for commenting! I get those testing trickles, and I can see they are essentially disappointed searchers where my content wasn't what they wanted. That's a downside to writing poorly targeted spoof material and on my list to go back over and improve.
Thanks Frannie - I can't resist funny, but I'm trying to get more on message.
whew! ok..I was quite worried there for one pixilated moment.
I don't think I've gotten any of the google + tick offs though I do get facebook likes. hmm. I wonder which is considered more important. probably the one I'm not getting...
Hi Flora, I like "tick offs". That's what they should have called it. The Google tick off.
Your content is so good that you won't need search engines as long as you keep writing.
Well thanks very much Paul! Until I get lucky / discovered I still need some of the basics in place. A bit like not giving up the day job just yet.
I think it is one of 500 signals which google uses to rank web pages.
Yes, well, 200 is the highest figure I've seen quoted.
However I am happy to only look at the one signal - did the reader find what they wanted on my page?
It saves a lot of discussion, theory and argument over the importance of the other 199.
Hey Mark, I have the google analytics too and looking at bounce rates, but I don't understand really what they mean (it's like math you know). Can you break that down for me? Say from one to 100? If the bounce rate is 100 percent what does that mean as far as what a reader is doing on the page? If it is 32 percent what does that mean? Help!!!!
Aaarrrggghhh! You're calling my bluff.
OK, I'll have a go.
100% means everybody who read it back arrowed. You won't get a time spent on page for a back arrow for some technical reason - although I would bet that Google Chrome can measure it.
0% means everybody who read it chose a different exit point. Hopefully they looked at one of your other pages, or exited via an ad placing.
In between - means a mixture of back arrows and good exits.
My 'best' measurement on a search page is just under 90%. Some Google guy said above 50% is dreadful.
But it depends on your page and the query. If your page answers the user then they go away happy. If you have a nicely linked set of pages that a user would naturally want to read more... then lucky you.
The lower, the better.
How's that?
This is great, thankyou so much Mark Ewbie, for taking the time, this really helps. I was just looking at all those 29 percents and thinking they were pretty low that I should be striving for the old 100 percent, just kidding, but kind of confusing when all you're doing is writing a poem about your cat...Cheers, snakeslane
Interesting announcement for Google today. We are going to get a lot less keyword data, so it's going to become harder to find out what terms are sending traffic to a page to clean them up.
-Paul (accidentally posted under wife's account:)
Good hub, Mark, with some great thoughts.
I'm not convinced that bounce rate is that important for all types of hubs, though. Mostly I write "how to" stuff and I expect the reader to come, find what they want, and leave. High bounce rate - near 100%
Hubs like this one, the reader might come and decide "that Mark guy is pretty smart - what else does he have?". Low bounce rate - near 0%.
Somewhere in between is reasonable, but I don't think it's anywhere near the 50% google says. More like 80% or higher for most informative hubs.
Repeat your last paragraph here - I'm no expert either.
I agree Wilderness - different content, different results. That 50% I only mentioned becuase it made my head spin.
I believe Google has extra info we don't. They list the search results - visitor chooses. Within a few seconds they are back looking at the results again.
Or within a few seconds they are away, done and answered.
A result I would assume Google can measure - did 'your' page satisfy the viewer or did they need more searches.
No matter how we slice and dice it though - for me certainly - ensuring that the viewer gets what they were looking for is (I think) the most important part of the 'game'.
Robin / Paul - that is interesting. I need to go look that up.
I looked it up. I think it says logged in Google users will be sending encrypted keywords for search privacy - these will show as a 'keyword' of (not provided) on your Analytic data. I had noticed these coming through and assumed it was some kind of spam phishing deal - bit of a relief to be honest.
It won't be for all searches by any means, so hopefully there will still be plenty of data to sit and pore over.
I just thought I left you a long comment. That sucks, it must have gotten deleted. Oh well, here goes again. I've been doing a ton of research on Google's latest algorithm updates, and everything I've been finding out backs up what you're saying, and what Paul has confirmed as well. Google seems to be using the user experience metrics as the overarching measure of quality. Gone seem the days of backlinking strategies and keyword stuffing.
I even agree with your analysis of all the traffic fluctuations and Serp fluctuations that have been happening a lot recently. I think they're spreading the traffic around to gather data and using the user metrics to decide where the Serps will finally settle in.
I've been cleaning house big time with all of my web content. I've deleted twice the number of hubs than I've kept. I want to make absolute sure that when Google shows up to my subdomain, all they find are high quality, on topic, well written, and well liked articles. I want to be considered an authority on my subject, with only high quality material.
I've actually been using the hubscore metric a lot recently. I've noticed that the hubscore for individual articles is actually a pretty good measure of overall quality and that the hubs with a lower hubscore either need to be heavily rewritten or else deleted. The hubscore seems to take into account view duration, traffic levels, and backlinks.
I hope your perseverance pays off. Hang in there. You're a great writer.
Hi Benjimester - thanks for having the perseverance and leaving a great comment... twice! The original was also there, although I haven't now approved it as it might lead to a case of duplicate content - I jest slightly nervously.
It's great to share opinions, particularly of course, if they are pretty much the same! I've been reading across the web, but the best 'advice' comes from your own experience - what is getting traffic, what is ranked, what needs to be done to clean out the stable.
I'm loath to delete anything, but I know it probably will need doing. For now I'm in rescuing mode, trying to make those unsuccessful pages better, and make the successful ones better too.
Thanks for the compliment, and good luck to you.
Mark great WAG, or is it? Any hoo i only write to pass the time. No one really takes the time to read my writing anyway.
Google Panda is that some strange bamboo eating teddy bear? Just having a bit of fun, make it a great day. Mike
Time will tell how much of a WAG it is. I'll pop by your hubs...
Thanks, your insightful humor is spot on by the way.
You were serious, for once, Mark, and I appreciate the points you made. I think you got it right.
Stephen Fry might say "the wish is father to the thought", but it's a nice idea that you are in control of your own destiny. Work hard and smart, and succeed.
I'll try it for a bit and see what happens...
I think a lot of what you say makes sense - however there may be an element missing- human frailty and ego. As a plunger I have seriously looked at what I write and have come to the conclusion that I write what the title says- however some of the keywords that used to find the titles were so remote from my title that I wonder how they got there. Like trying to find Newcastle by following signs to Bewcastle- the result is that the reader will be disappointed. The human factor is the one factor you cant allow for- if you produce a clear detailed article that is searched for but the reader actually wanted two or three lines to read without effort, they aint gonna stop but whizz on to the three liner idiots guide.- well that's my thoughts on the subject
Hi Case, please note I did not attempt to address the mysterious plunging issue.. and you guys have my full sympathy of course. I read somewhere, today, that just a few bad articles can screw a website - of course we still don't know for sure what bad is.
You are absolutely right about the reader - they may or may not like the style, content, layout or anything else. But I'm not greedy, they don't all have to like it.
Just some of them would be good I think.
I really hope you guys recover, figure it out. If I were in your position I might consider taking my "best" output, unpublishing it, and creating a new HP ID.
A bit like HP did when they gave us all subdomains.
Thanks again for starting a very worthwhile conversation here, filled with great humor, wisdom and advice. As for all the new SEO stuff, I'm saving my energy, trying to write better hubs and blogs...and you're spot on when you say, "One man’s favourite page is another man’s piece of rubbish." In any case, I agree with the others - you have created your own intelligent following, not judged by a robot, but real people.
Thanks GL. I like my HP readers and am very grateful. I just want some of that wide world search stuff as well.
snakeslane 7 months ago
Hi Mark Ewbie, this article is filled with useful content that I am absolutely sure people are googling for right now as we speak. I don't think I have any slaps pending...umm be right back, let me check my stats, Cheers, snakeslane