|Post Panda - New Ranking Factors|
We all know that Panda is supposed to increase the rankings for websites of higher quality, and lower the rankings for sites that are low quality. So what are these factors? What can google possibly be looking at to determine that one website is better than another?
Here are some factors I have seen or that have been mentioned somewhere along the way.
This seems to be a big one. Some people will say that ads play no role in panda but I am going to have to disagree with them. Ads can be a major part of your website, sometimes a little too major. If you can not tell where your ads end and your content begins, you may have too many ads that's causing your site to be punished by google.
The positioning of the ads can also play a factor. Giving advertisements prime position over your sites actual content is a sign that you would rather your visitors click on your ads instead of look at your content. Google can see this that your only purpose for your site is to direct your traffic to another source for monetary purposes.
2. Redundant/thin Content
If your site is about blue widgets, you might be inclined to put blue widgets everywhere possible to make sure google knows that your site is about blue widgets. Putting it in every page title or slightly varying the phrase to target other sub phrases can be detrimental to your efforts.
If you made a page just for the sole purpose of having it rank for a certain phrase and did not put much effort into the content on that page, you might be in trouble.
3. Lack of helpful information
Sometimes people do not know how to use your site. If they get lost on your site and have no way of getting help, this might be a negative signal to google. Putting a FAQ or a help section on your site can improve bounce rates and keep the user on your site longer.
4. Too much text / Not visually appealing
We all noticed the google site previews. So it is safe to say that they are looking at the site as a whole now and not just what text you have on a given page. Some negative factors here can be:
- Text is hard to read. Blends in with background too much or text is too small to red. Perhaps there are obstacles in the way that make it hard to read an entire article.
- Lack of videos, images and graphs to support your article. Rich media helps balance out any site. Some people learn better having something visual to look at such as images to further enhance what the article is talking about. Most typical MFA type sites are just page after page of text with little to no use of rich media objects.
- Text below the fold. Do you have a homepage full of rich content and then all the way at the bottom, a welcome paragraph that is there for the sole purpose of stuffing in important key words? If you need a welcome paragraph on your page, try to put it at the top, if its all the way at the bottom, it probably wont be read much anyway.
5. Offering something different
One thing I have noticed is that sites that stand out or offer something unique usually will be favored over dull, auto generated style pages.
For example, if you have an eCommerce site you probably have a lot of products that are listed on other ecommerce sites as well. Same product title, same product description etc. This is all fine, and contrary to what others believe, having the same product description as thousands of other websites will not get you in trouble with google. You will just be outranked by websites who are bringing more to the table. Some examples of this are:
- Linking to the user manual for the product.
- Showing reviews from members who have purchased the product.
- Showing products that are related or similar to the product that is being featured.
- Offering a unique review of the product, some things that might not be mentioned in the default product description.
- A video demonstration of the product in actual use.
- Unique images that show different angles of the product not found anywhere else.
I am personally tired of hearing the theory that using manufacturer descriptions will get you punished. That is not the case at all. Trust me, I have explored this theory over and over and over again. I have been building ecommerce websites for clients for over 14 years now and in 90% of the cases, we use the default product description. It is not wise to change the manufacturer description. Just like you would not change the title of the product, that would be bordering on misrepresentation of the product. It is not the duplicate description of the product that is getting you in trouble, its that you are not offering anything more than all the other ecommerce websites that sell the same product.
This is one of the reasons why Amazon ranks so high for almost every product they sell. They have unique user submitted images, videos, reviews etc of the products. They offer more unique data about a given product than most sites do. Want a great ecommerce site? Start studying what amazon does.
I planned on writing much more but I didn't realize how busy my day would get so I need to run for now. I hope others will feel free to contribute to the list in the mean time.
An excellent post, brinked - thank you!
What's interesting to me is the number of factors you mention that are about the ABSENCE of something, rather than the presence of something. I agree that this seems to be a factor, even though I certainly am not clear about how Google could assess it.
A great post and focus - very positive. You've captured my imagination and got the brain cells ticking.
Great to see a constructive start on a thread about Panda. Let's try to keep it that way folks and resist all temptation to descend into anti google diatribe!
My two penneth:-
1. Ads - don't have much experience. Was only running adsense on one small niche site 'til yesterday, when I added to main pandalised commerce site (1 block, foot of page). Not enough data for me to comment.
2. Redundant/thin content - I think redundant is the key. In recent history it was very profitable to keep stretching out that long tail, writing pages specifically for search keyphrases, even if the result was massive overlapping of page content across your site. IMHO, it is this that google now particularly despises.
3 & 4 Lack of helpful information & not visual appealing - for me these are combined and are causes of the same symptom that google monitors: the 'short click'. Google isn't going to keep sending people to your page/ site only to have them bounce back seconds later. Bad design and bad content/ lack of engaging content can both cause this. Trouble is, as webmasters we often can't see our own usability issues or how just plain wrong our site looks. In my experience, plenty of webmasters don't think this matters so much. I think first impression is everything. Get that wrong and good usability is for nothing and get that wrong and everything else you do is for nothing!
5. Offering something different - For me, key again. Google doesn't want to serve up 10 pages with the same pictures and same manufacturers description. Where's the value in that? Having something worthwhile that others don't is not only going to help differentiate you in the algo, but also assist with engagement (see 4 above) and achieving the long click.
6. Delivering what users expect - for me, this is the biggie. This is what Panda was all about. Google has put a lot of extra effort lately into trying to determine "searcher intent": what it is that you are trying to achieve beyond your three word query. We all know the basic types of searches, but I don't think previously google was very confident/ good at matching even to these. Too many articles showed up when I wanted to buy something! That's at the macro level, but IMHO it goes down to the micro level too. Gone are the days when you could get away with just writing about something (eHow, ezinearticles, etc made writing the cheapest currency in search). I think your page now needs to meet key factors specific to the search. It's a great way of separating the wheat from the chaff. For example, anyone can write about a travel destination, by regurgitating/ rewriting, but have you got any photos and are they unique? Having your own pictures kind of suggests you might actually have been there! Having library/ other people's pictures is better than nothing. Having no pictures is surely a warning signal?
That's my take on it as someone who has been pandalised and is now pretty much out of it (back on page 1 for most terms, ranked 5 - 10. Need a little more tweaking to get back in top 3).
Well, although all are good points, none of them is measured by Google directly, right? For instance, there are many valid reasons to have "text below the fold", I can';t imagine Google measuring that directly and penalizing for that.
After so many months into Panda, do we have a clue what factors are measured directly?
arikgub - the panda part of the algo is almost certainly a machine learning process attempting to get better and better at matching bad sites or good sites (depending on which way they've plumbed it). As such, it does not care for valid reasons, only that it has learned that users tend to prefer this/ dislike that for these types of searches/ this search. A degree of collateral damage is almost certainly accepted.
|Want a great ecommerce site? Start studying what amazon does. |
Surely, but then have a look at the bottom of their Best Selling Books page, keeping in mind:
|- Text below the fold. Do you have a homepage full of rich content and then all the way at the bottom, a welcome paragraph that is there for the sole purpose of stuffing in important key words? If you need a welcome paragraph on your page, try to put it at the top, if its all the way at the bottom, it probably wont be read much anyway. |
...for a good laugh :-)
I have agreed with point #3 since I realized Panda would not fix my issues. I've scoured my site with hours and hours of the "find and replace" function for similar content. Many pages were identified, perhaps three dozen of what I thought were minor issues out of 400. Four had major issues. Text was mostly eliminated or some pages just plain dropped as in the case of the latter.
This was early November. Not knowing whether or not a major Panda update has actually occurred since has me hanging. Has my site been combed through by Panda? Who knows?
Now I'm thinking major consolidation has to be done. What else could it be? This is the only issue I feel has affected my site (considering Google 23 points to look at if you think you were affected). I was years ahead of many with unique images and illustrations (above the fold to catch the users eye, but not overwhelming the content). I always felt text and just text was the most boring way of building pages. These are sites I always avoided in the past and still do today.
Back to major consolidation. I keep an eye on similar sites and where I used to rank. One site keeps coming back again and again, particularly one page that has 2,700 words! This is another type of design I don't spend much time looking for the information unless I use CTRL F. Most of my pages run in the 500 to 1,000 word range. I could probably hack (hate to use that word) the site down to 200 pages, but then it gets into that long page issue and the question of whether or not people will be patient enough for it.
The site is all about blue widgets...ah come on...but we can't drop keywords here. Yes it's a niche subject and some pages may look redundant but there are many ways to deal with a blue widget. Some are good, others bad, some can be used here but not there. Some are new, others old. They can be refurbished, bought locally or from overseas. The list goes on.
Frankly if I was to consolidate, the site would become what I feel is one that will create usability factors. Prior to Panda I had what I thought a decent page view per visitor count near 4.0. Now I'm lucky to see 2.4.
Which way to turn?
LostOne, if it helps, after trying everything myself since April 2011, the biggest impact has come from tightening up the focus of the site and reducing page overlap. Google is now much more apt to rank pages that don't even have the exact match, so you won't necessarily lose out here. If you don't want to delete, you can opt for noindex instead, which seems to be effective. For the record, we went from 30,000 pages to about 500, culling all the 'written for search engines content'. Simple test is, "would we have written this page if not to rank?"
Lets remember it's a moving target and "strategies" that are good today... may not help (or worse) in the future.
The ideas presented here are along the same vein that Google PR team has been saying for years:
|Just create awesome pages... and don't worry about us. |
Awesome web pages:
2. Competitive pricing
3. Compelling content
4. Additional images
6. (you fill in the rest for your niche)
Per the discussion of copying supplier descriptions:
I agree, totally not necessary to worry about it. With the caveat: If you copy the description and have 0 added value, Google will likely consider your site a thin/not-important web page.
Good points Brinked.
Regarding number 2. If your website isn't just about blue widgets, but is about blue widgets, green widgets and black widgets. Would you recommend still not using the keyword on pages specific to one of the widgets you support?
I was just taking a look at your website, which is specifically about blue widgets, and it's good to see that you practice what you preach.
Let me stress that these are not concrete known factors. These are not cut and dry. If you have text below the fold you wont necessarily be punished by google. I do not want anyone to think it is that simple. These things that I listed may or may not be ranking factors that are a part of an overall complex algorithm.
I am not claiming here to have found any secret ingredient that allows a website to suddenly break free of any google punishment. I have recovered many websites by improving the overall usefulness of the site in question. My approach is a "fix anything that could be looked at in a negative light by googles algorithm and then hope for the best". Kind of like firing a machine gun at a small target, you hope to fire enough bullets to get lucky enough to hit that small target.
The best approach when dealing with manufacturer descriptions is to look at your site without them. Google probably just ignores them anyway, so when that is gone, what else remains for google to look at and evaluate? If there is not much else, then you're site more than likely is going to be outranked by another business that sells the same item but offers more insight to help the customer make a more informed buying decision.
realmaverick, in regards to your question
"Regarding number 2. If your website isn't just about blue widgets, but is about blue widgets, green widgets and black widgets. Would you recommend still not using the keyword on pages specific to one of the widgets you support? "
If your website is truly about this diverse a collection of widgets, and I don;t blame you (widgets are awesome, right?). Then go ahead and feature those widgets prominently on your website, don't shy away.
Be more detailed about these widgets, don't cookie cutter yourself into a disaster zone. Have a page or section about green widgets? Dont just make your title tag "green widgets". That wont make me click on your page. What kind of information are you offering about those green widgets? A more useful title would be something like: "Learn all about green widgets" This tells me I can go to this page to learn about green widgets, since I am already on your site about blue widgets, I might be interested in green widgets as well.
If you pack that page with useful information about green widgets, that would be even better. Perhaps you did a study on green widgets that nobody else has done, that would surely get my attention. Ranking for "Green widgets" is ideal, I mean heck, thats what a lot of people search for right? We want our site to show on them high traffic searches right? Gone are the days where you can just insert that 2 word phrase and assume google will trust that your page was put up with the right intentions. Google now goes deeper and requires more than just a few keywords prominently placed.
Hope that makes sense.
Makes perfect sense.
I also agree, that there's no quick fix. The algo is extremely sensitive IMO.
You only have to look at Adwords, seemingly slight changes, can have a dramatic effect on your pages quality score.
I'd like to jump in here. My pandalized site was penalized on Feb 24th 2011. After several different types of changes, to no avail, I split the site into about 7 subdomains. Actually 6 subs that had highly "categorizable" content, and the original www kept whatever didn't really fit into the other 6.
The original (www.mysite.com) rebounded to almost original traffic, as well as a forum (forum.mysite.com). The other 5 stayed down.
I found an error on one of the 5 remaining subs that allowed several pages to reside under several unintended URL's. This was a database coding error that I finally figured out and fixed. That sub has returned to above pre-panda levels (up by about 10%).
The other 4 subs are still down 50-60%. I've consolidated thin content (#2 above), I've removed ads, I've searched for duplicated content in-site and out. I've fixed spelling/grammatical errors, broken links among many other things. I've been hoping that panda would redeem these last 4 subs since I made the changes back in early December. No luck. Surely panda would have done something by now if I had fixed the issues.
I am now wondering if I now have too much content/text (#4 above) on pages, and that they are visually unappealing since they don't have an abundance of pictures to support the text. So after reading what brinked has suggested, I feel I might try to be more concise on my long pages and try to come up with supporting images or rich media.
I'm getting closer, thank you all for your continued input and experiences.
|No luck. Surely panda would have done something by now if I had fixed the issues. |
Jan would probably be too soon. The upcoming refresh/update will likely be the test and it's looking more like Mar than Feb.
I always look to the business model. Even if you're doing everything "right", if you're not doing anything that sets you apart from your competitors the minute someone hits one of your pages, you're at risk.
For example, I would not go into the online shoe selling biz, because I would have a snowball's chance of standing out in that niche. And even if I spent tens of thousands of dollars on a shiny and well optimized website, Google would probably detect that I'm not really doing anything that a thousand other people aren't doing better.
Sometimes, before you change your website, you have to take a good hard objective look and consider whether you need to change your business model first.
Noticed you didn't discuss incoming links which according to some have a strong connection to Panda (and according to others they do not).
Now that I've had a year to do nearly anything internally that could be done to a site to recover, I've been spending more time on external factors like links. I find thousands of incoming links from junk blog sites. Being a long established site, these links have accumulated over 15 years.
In the past year my concentration has been in getting these off of sites with a larger number of links (say 100+) but now left with thousands below this threshold.
Is this worth my time to pursue for Panda recovery and how would I even go about it? The number is simply too many to be practical to handle.
<Google can see this that your only purpose for your site is to direct your traffic to another source for monetary purposes.>
isn't this what made them rich?
First they say above the fold, center page is best, now they punish us for it while they pepper ads all over the search pages. Theme targeted ads to boot, not near as many of the IBA's we get.
Simple test is, "would we have written this page if not to rank?"
I'll admit I started chasing the algo in 2010 with the addition of perhaps 50 pages, but they were not fluffy at all. It was actually in disgust of Ehow plowing me over with their regurgitated stuff. Maybe I should have stayed away from it. After all traffic has climbed steadily since inception of the site until that dreadful Feb 24, 2011 date.
I will probably take a deep look and see what was developed/added during that period. I'm also a believer that perhaps the site and many others are infected permanently and there's no way out.
Sure would be nice if the great Gorg could provide just one clue. Maybe as simple as showing whether or not Panda has gone through a site and a date? Then at least there is some kind of direction. Hard to ask for? Put it in WM tools