Google Algo Change - Human Behavior Theories

[browsee]

Just checked several losers from Feb 24th and Apr 11th, sites with dark/shady colors are affected. This is in line with their statement in Wired.

"We used our standard evaluation system that we've developed, where we basically sent out documents to outside testers. Then we asked the raters questions like: "Would you be comfortable giving this site your credit card? Would you be comfortable giving medicine prescribed by this site to your kids?"

So they tried to convert user behavior into an algorithm. I've been trying to understand how we can correlate human behavior to our website design.

Based from my research I've found that there is a relation between the Google algo change and Maslow's 'hierarchy of needs' theory in psychology.

I found this article describing Human Behavioral Theories That Can Be Applied to Web Design
[sixrevisions.com...]

Important factors in order of importance.

Accessibility: The website can be found and used by all people.
Stability: The website is consistent and trustworthy.
Usability: The website is user-friendly.
Reliability: The website is consistently available, without downtime.
Functionality: The website offers content, tools and services users value.
Flexibility: The website adapts to needs and wants of users.

Most of the Google engineers are PHDs, they tend to follow academics rather than reality. I am very confident that they included some of these behavioral theories in this algo update.

[Dan01]

That is some good stuff.

I did a little research and they were talking about this issue on this forum:

[webmasterworld.com...]

There were quite a few links and I haven't gone through it all yet. LOL I did see this:

Surprisingly, however, there is some indication that using white on black (or other light-text on dark-background combinations) is actually preferred by those with impaired vision...

Some say that too-high contrast is a problem for dyslexics, but too-low is a problem for everybody, and some other research suggests that dyslexics are fine with high-contrast text.

and then DrDoc said:

Technically white on black is easier to read than black on white.

HOWEVER ... since the web ultimately has its roots in the printing industry where everything is black on white (since no one wanted to waste ink by printing all but the letters) it is therefore naturally easier for people to read black on white, since that's what they are used to.

But then, there was Tedster:

However, most studies have shown that dark characters on a light background
are superior to light characters on a dark background (when the refresh rate is
fairly high). For example, Bauer and Cavonius (1980) found that participants
were 26% more accurate in reading text when they read it with dark characters
on a light background.

So apparently there have been conflicting studies.

[tedster]

I would lean toward studies that were done on a computer screen rather than on paper, and preferably on an LCD screen rather than an older CRT. I know the two I cited are on computer screens, but I'm not sure about the type - I'll try to dig them out.

However, if the color relationship of text/background is a direct factor in Panda, I'd be very surprised.

[Dan01]

I think there will be some eye fatigue associated reading lots of white text on a black background (light / dark). But for advertisers, they are not worried about fatigue. They want to make sure you see their ad and are able to read it.

However, if the color relationship of text/background is a direct factor in Panda, I'd be very surprised.

That is the important question.

[tedster]

You're not likely to get a straight answer on that kind of a direct question - see this quote from the TED conference interview about algorithm transparency, for why I say that:

Cutts: ...our most recent algorithm does contain signals that can be gamed. If that one were 100 percent transparent, the bad guys would know how to optimize their way back into the rankings.

Singhal:[ There is absolutely no algorithm out there which, when published, would not be gamed.

Cutts: I have to think, I have to hope, I have to aspire, there's some algorithm out there that we could publish as open source but couldn't be gamed. We haven't found it yet.

[wired.com...]
If you read that entire interview, you'll see that one single factor (like text color and background) cannot on its own be causing your ranking drop. The algorithm is essentially trying to detect the quality of the CONTENT.

My opinion is that you should let the text color idea go and look at other factors. Test a color change first, or just change it if you want - but I'd bet heavy odds that changing your colors will not restore your rankings. Some other combination of factors must be in play.

[Lexur]

I believe an engineer at Google proposed to insert a widget in the algo to measure content quality and nobody was able to tell him that was a stupid idea because no one really had tools to measure what he was talking about.

Then they left the widget run and it produced some good effects and some bad effects; it detected some scraper websites and ruined the life of some original content creators. Who cares?

Nobody can tell what are the factors of the new algo (the idea of the background color is simply amusing) and in Google nobody fears the back effects of this mess because, after all, actually there's no another search engine with his "quality".

So, altough I think I'm not affected directly, this Panda update is just and plan a piece of #*$! in the algorithm that nobody can move off now. I'd like to see some heads roll off in Mountain View but according Matt Cutts, "Panda is running as expected." Unfortunately, he can't define "expected" because nobody in the company can write that speech for him.

[Leosghost]

However, if the color relationship of text/background is a direct factor in Panda, I'd be very surprised.

As would I, the references to what studies the ad business has been using over the years is merely given to avoid people going down blind alleys ( awfull pun couldn't resist it ;-) ..and re-inventing the wheel .

Bear in mind too , it isn't because the ad business wants their ads to be visible and stand out that makes that a good thing on your site from the point of view of the user, there is another concurrent thread here where inspite of all the actual evidence it is still being mooted that intrusive ads and the ratio of ads to content in the initial viewport before scrolling is not the problem ..sites that look to the ordinary non webmaster user like they are "a page full of ads" or are "confusing to them", is not the entire problem, but IMO and IME and that of many many others who have not suffered with panda..it is a very large part of the problem of those who have...users do not like visual spam..and will not talk or post socially or recommend sites and pages that appear to them as visual spam.

Webmasters should watch non web master users ..rather than try to decide what their users prefer and what is or is not important in their users minds ..we are not our average users.

Those of us who have not been hit by panda don't need "proof"..about how users react to visually spammy sites compared to "clear" sites ..we already see it ..and are trying to pass knowledge on what works in panda and what doesn't, this knowledge we could simply keep to our selves,or even charge as clever consultants for it ( if we could stand dealing with webmasters for clients ..or stand clients in general ;-)), instead we are posting here, when we don't need to,( I have seen Huskypup also be told, that he does not "get it" ..and yet he has not been hit by panda..and Btw,like myself runs many sites with dark backgrounds ;-) however "one can lead a horse to water, but one cannot force it to drink" .

[browsee]

"one can lead a horse to water, but one cannot force it to drink"

Well said. But, in reality there is no water, algo created artificial water, you can't drink... so many horses are out of business.

[Leosghost]

The proportions of what makes up the constituent parts of the algo merely changed in relation to the past..

It has always been artificial, by definition machine intelligence is that, they have if anything folded in a more human flavour to the way the algo works, it now regards pages and sites more as a visitor would.

Those of us who made sites for human tastes and less obviously for Ad revenue have gained...I was not surprised in general at what went up and at what fell ..

People have always judged books by their covers..always judged things by quick glances ..always judged others by their faces..to think otherwise is to ignore biology and human nature and psychology..we all avoid the pushy salesman that tries to trick us or hustles us by thrusting his wares under our noses.

You catch more flies with honey than with...

Softly ,softly, catchee monkey ...

Busy, agitated, ad heavy pages are not "inviting" ..their message is "here be ads" ..not "here is something worth exploring or telling your friends about". and links from them are now worth less than they were, we all ignore the content of garish posters off line ..why would we believe they have merit and interest online.

Why would a search engine wish to send us to them ..its mission is not to provide the owners with jobs or income...

If your site looks like a fairground ..it will be treated as somewhere that one goes only for distraction..it must appeal and inspire confidence , if it looks like a clown wearing billboards and waving jumping banners in the wind..would you hang around it and ask for information..unless your intent was to actually visit the carnival or the circus..

Disclosure ..I've also produced many decors for carnivals and circuses..offline..nothing against them ..have many friends amongst their owners ..but they are not where you go for information.

[crobb305]

Two other factors occurred to me as possible signals (in addition to language detection and phrase-based document scoring, which are my main focus right now -- to improve my content)...

1) For all sites, compatibility issues could play a role (e.g., Browser compatibility problems? Display issues on mobile devices? Nested tables?) These could explain the uptick in well-ranking sites with clean code and well-designed templates.

2) Also...legal compliance (particularly important for e-commerce -- there are tools that will scan your site to test your legal compliance).

[Leosghost]

We are drifting into territories which are no longer "Human Behavior Theories" ..but , briefly ..

#1 ..no ..awful code ranks fine..( not that I have any ;-)
#2 ..no..legal requirements vary with jurisdiction ..a UK company is required by law to have its company registration number on its site..many don't ..they have not been downgraded in Panda.

You ( and many others ) need to think outside of what applies just to the USA .;-)

[crobb305]

We are drifting into territories which are no longer "Human Behavior Theories"...You ( and many others ) need to think outside of what applies just to the USA

I didn't say USA ;) But, Google has many people working around the world and they are aware of laws in every country (privacy laws, censorship regulations, defamation cases, i.e., Italy, etc); and yes, this is an issue that affects human behavior.

Before I do business on an e-commerce site, I make sure they have a valid and trusted Privacy Policy and that my transactions occur on SSL. This may or may not be a legal "compliance" issue in some countries, but it may be for others. My wording may have been poor, but these are ultimately legal issues. More importantly, they are TRUST issues. This item was the least of my concerns when brainstorming, and was just an idea. My main point was compatibility/display.

Yes...ugly versus clean code, nested tables, templates, and browser compatibility are 100% related to "human behavior". If the website doesn't load on my iPhone, I will not attempt to look at it. If the nested table loads slowly or doesn't display in my browser, I hit the back button. It's ALL ABOUT human behavior and relates completely to the discussions about pretty templates, images, black on white, text on road signs, etc.

I assume you read the entire thread, so you probably saw my post about the Microsoft website credibility study (fourth post of this thread). I am most considering human behavior, and have been from day 1.

But thank you for your guidance on "how to think" ;)

[edited by: crobb305 at 11:16 pm (utc) on Apr 15, 2011]

[Leosghost]

I meant

Also...legal compliance (particularly important for e-commerce -- there are tools that will scan your site to test your legal compliance).

"legal compliance"..
how could they rank 90% "compliance" in one jurisdiction against 89% "compliance" in another..what is legal in one is not legal in another ..they have diffrent teams around the world ..they do not however run "local compliance algos"..and then run them against each other.

And while I agree with you that clean code is better .."nested tables" take no longer to run in the majority of browsers used by the average surfer than "clean code"..the majority browser is still IE ..and it is far more forgiving even of missed tags than the others..

re .."read the entire thread" ..obviously I have ..I was replying to your specific post concerning compliance..because everyones countries compliance regs are different.. and you were assuming that non compliance with one specific countries rules ( USA ) could hurt sites to whom they do not apply.

Might as well say "spelling" would count ..which it can't, because in the eyes of all of us outside the USA ..you all spell incorrectly ;-)) ..and vice versa ..the spell checker doesn't realise that realise is not spelled realize .

Sure users get frustrated with slow loading pages ..and the more calls to off site ad servers and scripts and buttons and APIs that there are the worse it gets..( Googles adsense servers can be slower to respond than many ), that is certainly a valid reason and no doubt contributes to overall dissatisfaction and probably was included as factor in panda ..but nested tables are like a p-i-m-ple on a gnat in that scheme of things ..

It must be criteria that they can apply to all sites, everywhere ..because the low quality spammy content replicating scraping adfarms are not just all in the USA..

Chill OK..I'm trying to help ..on my time ..with no reason to do so ..not trying to tell you how to think ..just don't want you wasting your efforts and thinking energy on red herrings. ;-)

edited to get around the filter ;-)

[edited by: Leosghost at 11:26 pm (utc) on Apr 15, 2011]

[crobb305]

...just don't want you wasting your efforts and thinking energy on red herrings.

No red herrings here...at least I hope not. I have an advanced degree in a mathematical/physical science and I understand the principles of mathematical modeling; so, I am brainstorming factors that may influence humans' decision-making process and could be simulated/depicted by a numerical model.

the majority browser is still IE

Regarding display and nested tables, I see your point about browsers, but my concern is connection speed -- many, many people still use slow connections, and mobile compatibility is becoming increasingly important. One of my sites soared during Panda. It is in a beautiful template and loads in half the time of my Pandalized site, which is very old and was built with nested tables. All things equal, both sites are identical. They even have the same number of affiliate links (not many, just 2 or 3 across a couple of pages). The only difference I can find is the html/template.

[edited by: crobb305 at 12:18 am (utc) on Apr 16, 2011]

[Leosghost]

lets try to keep it friendly huh ;-) ..by "my own time"..I mean I didn't get hit by panda ..so to me it makes what you would say is "no nevermind"? ;-) if some recover ..or not ..But you are actually one of the guys whom I'm trying to help ..which is why we have exchanged stickies...yeah ?..most who haven't been hit are not posting on the subject..or have given up posting on it ..because they get shouted down..this is IMO one of the better threads on the subject ..
Because you and others are actually using your brains ..and not just blaming Google or in denial about ads playing a large part..

I'm here posting at 01.40am my time and still haven't eaten dinner ( actually I still haven't cooked mine yet ;-)( and since panda this has been the case many times for us )..my wife tells me I'm crazy as we didn't get hit so why am I trying to help out folks, who then rear up at me for doing so ? ..as she pointed out yesterday .."funny how if someone has admin or mod by their name and they say the same things I do" ..no one jumps down their throats..or accuses them of "schooling"..
If a member who signed up yesterday told me where I was going wrong in perl.. rather than sat back and watched me making syntax errors ..I'd be a happy guy.;-)

[Leosghost]

many, many people still use slow connections, and mobile compatibility is becoming increasingly important

Amen to that ;-) ..and especially
.

many people still use slow connections

I keep one slow machine connected to the net just for that reason ..it locks up and has me and the family looking for something to throw at it on some pages ..and yet it has 500 megs of RAM and runs a Pentium 4 and Opera. Linux box ..Mint7

but more than 4 tabs today on a french price comparison site and one tab on Ebay ( my wife was "window shopping" ) and it locked ..froze solid..

I quote
" my mouse isn't working ..the bar at the side wont move ..it wont click" ..
mouse was fine..
all the ram and the CPU cycles were gone with flash animated ads running in the header of the price site.

( and they were not big ads ..but they were making the machine like molasses ..because each tab ..even those not "in focus" were running them)

..and I hate to think what ebay was doing in its tab ..

I had to do a hard reboot and tell "her only use the same tab on some sites ..don't let them open new ones."..She is very much what you call an average user ..intelligent but not at all geek ..not interested in all that ..

Her solution ..quote ( and translated from the French ;-) "OK ..when it starts again ..I won't go to that site again ..not if I have to remember to do that ..its all too complicated , why cant they get it right if you can"..multiply by a few hundred thousand similar experiences ..and Voila !

The kind of site that G just zapped with panda .and her behaviour is typical ...the rest of their code is pretty clean ..I opened it on one of my machines and checked the source.

But their designer forgot ..not everyone has his machine ..or the ultra fast laptop his boss does..most of their clients probably have one like she used.

[crobb305]

I'm here posting at 01.40am my time and still haven't eaten dinner

You need to eat! I do too actually. I drink too many energy drinks and I type way too quickly. I can sound abrasive without even realising (realizing?) lol. I edited my last post down a bit to sound friendlier and to clarify my reasoning a bit more.

I have been trying to take a step back and think of things that I look for when making a purchase online or when determining if the article I am looking at is credible and worth printing/sharing...things that affect my decision making process. Granted, these factors are really in addition to "content quality" Google mentioned, which is a separate issue I think.

BTW, it didn't occur to me initially that you were the friendly ghost who emailed me the other day. That was much appreciated :)

[Leosghost]

If I could type faster and more accurately in English on this French keyboard without hammering the keys and waking the house, I could edit better , and sound as friendly as I am sitting here;-) ..anyway no harm done..;-)

Before I eat ..food for thought ..you may have hit a very important point

with mobile compatibility is becoming increasingly important

Given Google's increasing shift into mobile and android on portable devices and into "cloud" and away from desktops running an OS ( at least thats where they'd like end users to go )..maybe a good mobile version was a criteria ? ..or at least a site that would fit on a mobile or a tablet screen without any/lots of side scroll?

I always tended to build "liquid" and with "white space" so my sites can compress into "mobile" OK.

A friend here told me how he went to "fixed" because he saw his "liquid" designs one day on a 32" screen and "they just looked silly" ( his phrase ) ..Got me thinking about going fixed ..when I get the time ..

Lot of other considerations on mobile too ..data caps ..here we have none ..you guys do I think ? ..here we already have browsers and web and webmail/ email on TV in 1080HD ..I just got delivered my new routers and tivo type but better TV box in the same pack.. $50.oo per month free international phone calls to fixed and cells and no data caps at all ..( the kind of thing that Steve Jobs doesn't think will catch on ;-) and included is free streaming movies and built in games console and blueray .

If Google are getting into TV and web via TV ..they may well be looking at how sites would fit when that goes mainstream in other countries ..don't know if they'd give the US version without data caps though ..I use terrabytes(sp?) by day every day even without the TV service...and my son uses You tube in HD heavily and frequents animation and 3D and flash fora ..he works in that stuff..lots of data moving ".

[crobb305]

Given Google's increasing shift into mobile and android on portable devices and into "cloud" and away from desktops running an OS ( at least thats where they'd like end users to go )..maybe a good mobile version was a criteria ? ..or at least a site that would fit on a mobile or a tablet screen without any/lots of side scroll?

Yeah, the mobile compatibility was what led me on this path today. I tried to view my Pandalized site from my phone and it wasn't displaying properly and scrolled endlessly. I ran my site through a compatibility tester and it failed miserably for mobile compatibility (at least according to the tool's assessment). It passed for most other browsers. Not a lot of images, but it was built 7 years ago using nested tables. When I am on a very slow connection (e.g., using a router that's on the other side of the house), I have observed my site gradually load into the browser one table cell at a time. So I imagine the tables (especially when nested) can slow things down.

I just find it interesting that my other very similar site soared with Panda, displays well on my phone, passes all compatibility tests, and was built in a well-constructed template. Of course, if the template is playing a role in Panda, it could all come down the "human factor" (assuming templates just get additional credibility). I personally think it goes deeper than that. We're talking about a complex algorithm that took a year to construct. I think the source code is partially implicated, with considerations for the "web of tomorrow".

[edited by: crobb305 at 1:21 am (utc) on Apr 16, 2011]

[Leosghost]

Not a lot of images, but it is built in nested tables. When on a very slow connection (e.g., using a router that's on the other side of the house), I have observed my site gradually pop in one table cell at a time, so I imagine the tables (especially when nested) can slow things down.

Ahah! ..Now there I totally withdraw my previous comment re nested tables ;-)) My apologies are freely given I was fixated on nested tables in desktop browsers...had'nt thought about them at all in relation to mobile device browsers.

Wonder how any of G's "testers" were using mobile devices via wi-fi ..must have been some ..given they are betting the farm on mobile ..it isn't the entire story ..like ads '(even dancing ones ) aren't the entire story ) or "social engagement" but I'll bet they are all three of them at least pieces of it.

This is why I like threads that "drift".;-)).linear thinking or separate threads for each idea get in the way of "association" and "conversation"..even if they get "lively" ;-)) as long as no one is being dogmatic about what it absolutely most definitely cannot be, and some real detail argument is used instead of exaggeration or inaccurate generalisations ( spell checker wants me to put a "z" in there again ..really slows me up when the box "bugs me" about what "it" thinks is my spelling errors ) or accusations of "500 people all saying".. we all learn ..its good ;-)

[crobb305]

This is why I like threads that "drift"

hehe. Yeah, my initial reasoning had human behavior in mind, even if the underlying causes & solutions are entirely different discussions altogether. :)

[bramley]

I have an idea that pertains to this thread and how the algo could / should evolve that will appear in a new thread. Been waiting 5 hours now and need to sleep ...

< here's a link to the new thread: [webmasterworld.com...] >

[edited by: tedster at 2:19 pm (utc) on Apr 16, 2011]

[Dan01]

Leo:

Busy, agitated, ad heavy pages are not "inviting" ..their message is "here be ads" ..not "here is something worth exploring or telling your friends about". and links from them are now worth less than they were, we all ignore the content of garish posters off line ..why would we believe they have merit and interest online.

I agree. When the Adsense rep called me, they wanted me to change my top ad above the content to a 300X250 image / text ad. Way too much, IMO. I didn't want to do it.

Also, your comment reminds me of those flashing bright red / green ads. They really trashed up the sites they were on. Maybe it was Adbrite that had them.

I see a lot of webmasters place two 300X250 Adsense blocks on their website, pushing their content below the fold. This has gotta ruin the user experience.

Another big mistake is having a huge picture at pushing their text into a one-inch wide column. Again, it makes the site look trashy. On one website I had a right and left side bar. This made the content only three inches (maybe 350 pixels?) wide. I had to widen the whole thing so that if a picture was inserted, it looked right.

One more thing. I was on a major website today. They had so many ads in the right side-bar that the content stopped, but the right side bar pushed the footer way down the page. It looked funky and not very professional IMO.

[Seb7]

I think trustworthiness is very hard to evaluate, as dodgy webmasters seem to know all the tricks, they know how to make a website look authentic. Ive seen some really good websites which obviously had a lot of time and work gone in to it, the only reason I knew it was dodgy was how I arrived at the page. On the flipside there are millions of websites create by trusted people but had little design knowledge, thus it didnt look too good, and could easily be classed as dodgy.

All my sites do seem to be performing better than ever in the SERPs now. Multiple pages from the same domain on the first page becoming quite common.

I just wish google would work on the long tail keywords, as this seem a big fail lately. Whenever I'm searching with more than about 4 words, the results are petty rubbish.

[Leosghost]

Brinked has made some very good observations in another thread ..but is getting the same kind of stubborn denial of "it can't be"..that keeps me out of that conversation.

[potentialgeek]

I'd recommend every major site that got hit by Panda consult with a usability firm which does website user testing. Google has said that 84% of the hit sites had already been blocked by website visitors (using Chrome). Unless you know with 100% certainty you were among the 16% which didn't get blocked, the key to reversing the financial loss could be usability testing. It can be extremely difficult to be objective without fresh eyes on the web design you've stared at a thousand times. Even if the usability tests don't break you out of the Panda punishment, they should at least pave the way for your site to get more quality links from satisfied users.

[browsee]

@potentialgeek, usability testing is a nice idea. I think demographics play a major role here. Alexa normally gives demographic information. Probably they are targeting sites based on demographics. Every site is different. It is not just one factor but a combination of several factors.

[tedster]

Probably they are targeting sites based on demographics

That's an interesting human factors idea. I've already thought that Panda had to handle certain "silos" of websites differently. In other words, e-commerce might get a differently weighted mix of factors than news or articles. Some factors might even apply only in one silo and not in others.

I'd bet that Google has pretty good demographic data about site traffic, so they certainly might be looking at that for another way to slice things.

[Sgt_Kickaxe]

Facts, we don't have any besides an admitted statement that Google is trying an algorithmic approach to mimic human judgment of trust.

It's clear that there is human review going on for pages that reach page one rankings though an algorithm is likely flagging such sites for human review. Making your site ready for such a human sniff test may be the only thing to worry about, imo. It's also possible that if a human reviewer looks at 5-10 pages and finds only questionable content the entire site may be "adjusted"

Proper title tags and H1 tag use and social buttons and, and, and... none of it can help if you're trying to pass off low quality content, or MFA site, as legit via ANY code methods since a human will undoubtedly make a quality decision instantly.

[walkman]

Sgt_Kickaxe, sad that one split second from a googler having a bad day and your livelyhood may be gone. No warnings and obviously no appeals. We need 4-5 search engines

[tedster]

sad that one split second from a googler having a bad day and your livelyhood may be gone

From what I've read in the patents and interviews with Google spokespeople, one human reviewer does not have that kind of power. They do provide input, but that input gets aggregated and passed on to a supervisory level. For example, see Google Patent - human editorial opinion [webmasterworld.com]

But I certainly agree with you - we need more than one strong search engine. Yandex, Baidu and Bing are not serious Google rivals for most of the world.