Welcome to the Q&A Forum
Browse the forum for helpful insights and fresh discussions about all things SEO.
I'm with Ash on the Internationalization strategy. I would also suggest that if you go with domain.com/countryspecificsection/ then each country specific section should have it's on country / language specific structured markup and Meta data assignment. This will help ensure Google doesn't have to figure it out on their own (because they are a gambler's nightmare as to how they can get mixed signals wrong).
There could be many reasons the site has lost rankings.
Does Google Webmaster Tools report that the site has been manually penalized for bad links? (check both the "site messages" section and the "Search Traffic / Manual Actions" section in GWT).
What about other factors like on-site problems such as topical dilution? When I view the Residential Elevators page there's a lot of content that has nothing to do with "residential elevators", let alone that the company has experience installing residential elevators in Wisconsin or Illinois. For example all the "Recent News" sidebar content, and all the "Our Product Categories" content on the lower portion of the page severely dilutes the unique content of the page. As a result, there's likely not enough depth for the primary focus of the page.
This is made worse by the fact that that sidebar and lower portion content is duplicated across the site. The end result for just this one factor is that important pages are likely being considered "thin" overall.
Another major problem that could be contributing to the loss of rankings is site page processing speed. My test of the home page showed fourteen seconds - Google has the ideal page processing speed for an individual page at between one and three seconds. Such slow processing harms overall trust and quality.
There could be a dozen other issues impacting the site's rankings as well, such as , however those were the first few I found that all need addressing.
Harald,
I need to challenge your response here.
You stated:
"Boost works not only for manual actions but also for algorithmic problems, like Penguin."
That's absolutely not possible. Penguin is not baked into the main algorithmic process. it's a separate entity that only runs a couple times a year. So unless the clean-up was done literally within a few days of the next Penguin update, the magic recovery you claim was NOT due to Penguin related problems.
Penguin is a penalty that was created to flag unnatural, artificial linking where that linking was implemented purely to game the Google system.
If both site A and site B are legitimate, trustworthy sites, where the company brand is clearly consistent across both sites, and as long as the links are a "buy" link that takes the visitor to a purchase page, and where those links have anchor text that is not stuffed with keywords, then there is no reason to believe Penguin would flag either site for artificial manipulative links.
And yes, putting a nofollow on those links will ensure this is an absolute certainty.
Yeah it looks like you just have a lot of work ahead of you to boost your ranking signals.
When I did a Google search for "DPF cleaning commercial remapping" (not using the quotes), you do show up in the #4 position organically. So Google knows your site is there. And it has 8 pages indexed.
So the questions to ask include:
Do you have enough overall content across your entire site to properly compete against other sites that might have extensive content specific to your targeted market?
Even if the existing site's content gets a top grade as it stands from basic SEO signals, how strong is each page itself, in comparison to any other pages on other sites specific to that single page's primary topical focus?
How strong is your off-site signal footprint compared to those competitors?
I agree with @adamxj2
Parallax DOES cause problems for SEO when you're only serving one URL that can be optimized. In your situation, you're running independent URLs, and going directly to them you can see at the source code level that the page Title is unique to that page.
Because of that, I am pretty confident that as @adamxj2 said, it's an issue attributable to the separate issue of where Google changes page Titles and meta Descriptions that are being displayed based on their algorithm's assessment and the belief that method is better for the searcher.
Business listing directory environments have a big challenge when it comes to URL structure / information architecture and content organization because:
So what is a site to do?
The best path is to recognize that as mobile becomes more and more critical to searcher needs, hyper-local optimization becomes more critical. It becomes the most important focus for SEO.
As a result, URL structure needs to reflect hyper-local first and foremost. So:
This way, if someone searches for "Pizza Hut Dehli", all of the Dehli Pizza Huts will show up, regardless of neighborhood, while anyone searching for "Pizza Hut Saket" will get more micro-locally relevant results.
And for those businesses that serve a wider geo-area, even though they too will be assigned a hyper-local final destination page, they will still be related to their broader geo-area as well. So someone searching "plumbers in Dehli" will get the right results and then they can choose any of the plumbers in Dehli regardless of what neighborhood they are in.
Note how I removed /search/ from the URL structure as well. It's an irrelevant level.
having the clearer understanding about the concept of having multiple "titles" on a single page (an H1 headline is the in-content "title" for that page), David is correct - while HTML 5 allows multiple H1 tags on a single page, this is bad because the H1 communicates "this is the primary topical focus of this unique page".
Because of that, if you have headlines within the content area for content elsewhere on the site, and link to that other content, then those are absolutely best served with H2 headline tags, or if not , then at the very least, "strong" tags if the topic of each target page is significantly different than the primary topic of the page they're all listed on.
Kingalan1
I'm not a programmer by trade - the way I begin even considering these things is by running tests on various tool platforms.
For example, put a page you think is slow into URIValet.com - test as Googlebot. The resulting report has a block of information in it regarding total size of files processed. It breaks that data down to file types. Look at the CSS/JS lines - if they are more than 50k to 100k total for either CSS or JSS, there is almost certainly inefficiency in there, and likely unnecessary bloat.
Go to WebPageTest.org and do the same - put in the URL you want to check - choose a server location and DSL (which gives a fair mid-range speed evaluation), and Chrome as the browser emulator. The resulting report gives you a lot of information, however the one page in that report that may be most helpful in this situation is the "Details" report - if you go there, and scroll down, you'll get to the section that lists, line by line, every single file, script, image and asset processed for that page, and all of the data on speed of processing each step of the way (such as First Byte Time, DNS lookup, SSL lookup, and more). Those can reveal several individual bottleneck points.
Damon,
To start, let's be clear - Panda isn't a "penalty" - it's an algorithmic adjustment based on quality, uniqueness, relevance and trust signals.
Having audited many sites hit by the range of Panda updates, I have a pretty good understanding of what it usually takes. so having said that, I took a quick look at the site. While Andy may be correct in that you may only need to wait and hope the next or some future Panda update acknowledges the changes you've made to this point, that very well may not be enough.
1st obvious problem - your site's response times are toxic. - a crawl using Screaming Frog shows many of the pages have a response time of between 3 and 7 seconds. That's a major red flag - response times are the amount of time it takes to get to each URL. If it takes more than 2 seconds, that's typically an indicator that crawl efficiency is very weak. Crawl efficiency is a cornerstone of Panda because it reflects what is almost certainly a larger overall page processing time problem. Since Google sets a standard "ideal" page processing time of between one and three seconds, if it takes more than that just to ping the URL, the total processing time is likely going to be significantly worse.
While it's not required to always get a one to three second total process time, if too many pages are too slow across enough connection types for your visitors, that will definitely harm your site from a quality perspective.
And if too many pages have severely slow response times, Google will often abandon site crawl, which is another problem.
Next, I checked Google Page Speed Insights. Your home page scored a dismal 68 out of a possible 100 points for desktop users (85 is generally considered a good passing grade). That reinforces my concern about crawl inefficiency and poor page processing. It was even worse for mobile - scoring only 53 out of 100 points. In my second test, I got 63/100 for desktop and 49 for mobile. The different results for the two tests is due to the fact that speeds are worse at different times than others.
Just one of the issues GPSI lists is server response time (which confirms the very poor response times I saw in Screaming Frog).
Next, a partial crawl using Screaming Frog crawled 20 URLs that resulted in 404 (not found) status, which means you have internal links on your site pointing to dead ends - another quality hit. And SF found 25 internal URLs that redirect via 301 - further reinforcing crawl inefficiency. Since this was a partial crawl, those problems could be even bigger scaled across the site.
Then I poked around the site itself. http://www.workingvoices.com/courses/presentation-skills-training/keynote-speaker/ is indexed in google, as it's one of your courses. That page is possibly problematic due to the fact that there is hardly any content on that page overall. So while you may think you've dealt with thin content already, I don't think you fully grasp the need for strong, robust depth of content specific to each topic you consider important.
That's nowhere near a full audit, however the above are all examples of issues that absolutely relate to working toward a highly trusted site from Google's algorithmic perspective.
Filolari
I just tried manually - and again with personalized search turned off. In both cases, the site was nowhere to be found in the first eight pages of results organically. I tried with my location settings set to my location (Corte Madera, CA), and location set to United States (which eliminates local search bias to a specific geographic area within the U.S.).
I'll build on what Mike said in that comment links are less valuable than they used to be due to the proliferation of this as a spam tactic. And I expect they'll become even less valued as time marches on. For the most part the value is weak at best nowadays, and requires a lot more time and effort to get right given the attention this type of link has garnered from Google due to the spam issue.
I'm in wait-see mode. I expect it will get some traction, yet think it's a terrible idea - setting a bad precedent - it automatically penalizes sites that don't have it. Not good at all. So for now, I wait it out in protest.
yes, part of it could be dynamic sitemaps, however consider these factors:
1. According to their Moz ranking, Yelp has a Domain Authority of 93/100 - that's a lot of domain authority.
2. They've got over 16 million pages indexed at Google.
3. Of these, about 5 million are search results
4. Google claims to be able to explore forms - their bots are claimed to be able to simulate human behavior in form usage.
5. With about ten million pages indexed that are not yelp search results pages, that's an awful lot of raw data for Google's bots to experiment with in trying out that search for.
akitmane
Unfortunately there's not publicly available data on exact counts of new content or back links being a magic number. Every site in every market is different. What I can offer is this - first, if every article you write is high quality and unique, you could write many articles a day and should be fine. Each article would be weighed on its own and the entire site as a single entity at each Google pass.
As for using an automated link generator, that's where I'd be very cautious. Have you confirmed that the links you get from that source are coming from related, relevant sites? Are they diverse links in terms of anchor text? Are the links you get over time coming from a wide variety of domains located across a variety of I.P. blocks? Automated link building solutions are usually very risky at best. Especially if you're not obtaining links through means other than that one source.
Links should come from a wide variety of sources - and if a pattern of relationships can be detected it's a red flag. The actual number of links you get from one single natural path, such as submitting a viral piece of content could literally total in the thousands all in one day. Yet I highly doubt any automated link generating solution could match the diversity that would result in, though it's possible such software exists.
A solution like the one you mention is very questionable just from the marketing used on that single page site. Words like "how to legally hack", and "top secret", the fact that it's a one page site, just about EVERYTHING in that content screams it's most likely not a long term trustworthy solution.
Well it makes sense that an Italian domain would show up in the mix for you when you search from a location in Italy, but not when searching from within the U.S. - there are many U.S. based sites that contain that phrase. And just because you clear out your cache, etc., be aware that Google still presents some degree of personalization based on country.
Mario,
No offense, but you apparently don't understand SEO best practices.
Example 1: The links in the footer area pointing to each town? That box with all those links is an old-school method of driving SEO value. Nowadays, links in footer areas offer much less ranking weight. When they're all stuffed with the same words, that's even worse because it pretty much waves a red flag saying "spam over here".
Example 2: Links in the footer for various towns - there is so much duplication of content across these pages that they're all harming you, not helping. When a page has mostly the same content as another page, it's a problem. Multiply that by every town linked, it's a nightmare.
Example 3: when I go to one of those and click into the DUI & Criminal Defense link, it doesn't go to a unique services details page with several paragraphs of unique text. Instead, it goes to a page with a small amount of text followed by several links. So pages at this level are nothing more than link pages, and thus offer little value to the overall SEO.
Example 4: http://www.musillawfirm.com/criminal-defense/criminal-mischief/ Look at that page. There's almost no unique content dedicated to that page's topic. What little text is there that's unique is overwhelmed by the duplicate content that exists in the header, sidebar and footer area). Why would search engines consider this page important to someone doing a search?
So before you do any more "optimization", I would highly encourage you to learn what real SEO in 2011 means.
I don't use ping.fm - I'm talking purely for maintaining a twitter stream. I'm a power tweeter. Or more aptly, addict. Twitter is the single most important resource I have for staying on top of cutting edge industry trends, topics and debates. And it's how I maintain my own voice being heard above the chatter.
I have a column in TweetDeck set up where I have assigned the most important "pay attention to" people out of all the people I follow. That way their tweets don't get lost in the mix. And from time to time, I scan the longer "everyone" column.
Like I said. He's doing some right things, although by having them be dofollow, that's in direct opposition to Google's saying they should be nofollowed.
And just because "other sites get away with it" does not make for an intelligent, wise or smart reason to do something. Roll the dice.
Whether you include the sub-nav in the site-wide navigation as drop-downs or not, make sure that you also have that sub-nav as on-page links (such as in a side-bar) specific to each section. So that only that sections page links will then be on all of the pages just within that section, without the drop-down needing to be used. This ensures proper attribution and relationship queues for SEO.
Having said that, this issue is widely debated in the industry. I've seen arguments on both sides claiming they have proof that one or the other of your choices was better. Personally, I've always taken the approach that SEO should always be balanced with usability. So when this question comes up I don't tell clients to kill the site-wide drop-downs. Instead, I focus on other refinement (like the section by section dedicated side-bar links, and proper breadcrumbs, and in-content links, and so forth). And that has always gotten my clients solid results.