I am puzzled by this site's rankings

Joined
Aug 15, 2015
Messages
53
Likes
34
Degree
0
Hey guys, I am SEOing this site for the past 8-9 months, it behaves very strangely and honestly I am bit confused and need some ideas to nudge it into the right path.

It is a site of a local business in big US city. The domain is an exact match, even though according to most of the KW tools the key phrase has something like 0-10 monthly searches. The competition is low to moderate. Top competitors have between 250-450 referring domains, some of them are legit magazines mentions, links from local bloggers, business citations/links from various chambers of commerce, some profile links, blog comments and forum links. Anchor wise compared to them I believe that I am in the norm, as far as I can judge from their public links.

About the site link profile. The interesting thing is that I managed to rank the site for all the major KWs (around 15-20) between positions 3-8 with a bunch of local citations, 40-50 profile/IMG/video links, 1-2 web 2.0s, and some cheap PBNs on T2 pointing to some of the profile links and the structured citations. The guys who worked on the sites before me had built 4-5 exact match contextual links from link farms (nothing special about these sites) and not more than 25-30 local citations. In November all of the primary KWs popped on page 1 overnight, that was very strange to me, to say the least. The site stood on page 1 for a month, in the middle of December, when there were unconfirmed G update and some big fluctuations, most of the major KWs dropped 10-20 positions overnight.

For the past 5 months, the rankings are fluctuating between page 2 and 5. In December when the drop happened, I thought it was an on-page problem, caused because of the domain name, and general KW density. Just to mention all the main KWs are ranking on the homepage. Also, all of the menus was containing the primary KW etc. Generally, I lowered the KW density. The rankings dropped even furthers, there were some KWs still on page 1, but after the change, they dropped as well. Later on, I came to conclusion that the drop was caused because my current links were devaluated, ofc not pretty sure about that, just a theory.

Off page wise, since December I've built,

- 40 paid PBNs (Anchors: mostly generic, naked URLs, 2-3 exact and 2-3 partial match).
- around 15 web 2.0s backed with cheap PBNs (Anchors: naked URLs and generic)
- 40-50 structured citations again with PBNs on T2
- some profile links and a press release (used naked URL as anchor), few info grpahic submissions and other pillow links.

When links are built it's like thy are not influencing the rankings. I can't say the links are great but still. Last month I bought 15 PBNs, made the guy drop 5 of them with KW rich anchors, the rankings improved to page 2. The next 10 links from the package were with 7 naked URLs anchors, and 2 of them are LSIs. Since then the rankings dropped like never before.

I know it's hard for you to give advice based only on few sentences but still, but what's your gut feeling about the whole situation?

I have a feeling that I should be more aggressive with the anchors and distribute more links to some of the blog articles and link back to the home from there, may be add more KW rich internal links. Also, at the same time, I am wondering if I tripped a filter somewhere along the way and G is playing me.

Here is a screenshot of the primary KWs rankings:
hUaEzTb.jpg
 
I know a lot of people still have great success in local with their 'real PBNs' (ie ones they built themselves and don't overuse/put at risk from public sales). 'Cheap PBN' probably = most links known to Google and of very low value/fully devalued. Either way I hope you've been ethical and fully disclosed the risk to the business owner (I'm sure you have if you're going so aggressive on the SEO and he's hired people in the past doing similar stuff...).

Local is kinda easy. Except in highly competitive niches in cities with a billion contractors (like Philly, New York etc etc) you can rank with a handful of natural links, doing the citations right, and making sure the content is spot on, and getting some good reviews to help boost the map listing and performance.

You've done the citations so you know the problem, realistically, can only lie in a couple of remaining areas:

Reviews
On-site content
Google doesn't value the links you have

If you think you have the other pegs in place then you probably want to revisit the links and either use better PBNs (if that's the strategy you and business owner have agreed - presumably to minimize the budget) or better still future proof things by doing some linkbait, natural outreach (you really won't have to do tons... 10 really good natural links might do the job if you're already on page 2). You can generally 'phase out' pbn links by taking them down slowly - replacing 1 for 1 with natural links and still move up as the natural links tend to be more powerful.

You're much less likely to be back in a year wondering why you're back on page 2 again if you turn him towards a more natural strategy, and obviously you can charge more for your time if you're doing natural PR-style outreach promoting his content.
 
Let me tell you the three main reasons for what you're seeing, in my opinion:

1) Changing a rank of a document by applying a rank transition function

"In December when the drop happened... I lowered the KW density. The rankings dropped even furthers, there were some KWs still on page 1, but after the change, they dropped as well."
That's what you said. Here is what the patent says:

The rank transition function provides confusing indications of the impact on rank in response to rank-modifying spamming activities. The systems and methods may also observe spammers’ reactions to rank changes caused by the rank transition function to identify documents that are actively being manipulated. This assists in the identification of rank-modifying spammers.

Your site is a perfect example of this patent at work. It's actively been in place for probably a decade now. So basically, when you make SEO-focused changes to a page (links or on-page), Google will do one of three things:
  • Nothing
  • Positive Change
  • Negative Change
All three indicate that you've done something to increase your rankings and you should be celebrating at seeing any "bouncing" or "dancing." It's a good sign... for those who know about this.

For those that don't, they react. Non-SEO's don't react to anything SEO related. They look at macro-trends, if even that. And the reaction during this period is exactly what Google is looking for, because that is you exposing yourself as a spammer.

This period of "randomized" movement can last for 0 days if you have authority and trust, or up to as much as about 90 days.

If you start tinkering during that time period, you're toast. Tinkering to test Google's reaction used to be possible, and they made this algo-patent to put an end to it. And those that keep trying are painting a target on their back. They want to destroy anyone trying to reverse-engineer their algorithm.

This is consistent with the things you go on to say about not seeing anything positive occur, and then the slow decline in your graphs as you continue to react instead of allowing it to stabilize.

2) Google is back to allowing bad links to give negative page rank

This is from my own testing. Twice now, I've disavowed tons of spam on two sites. One I know was penalized from an early Penguin. I got it unpenalized but it still felt throttled so I sought out more spam to disavow, and did, and watched it finally get "unlocked."

On another site, I knew I should be seeing far more results than I was getting for each campaign. I recognized the throttling, disavowed some direct spam I had done in very small amounts (which used to be fine), and I immediately unlocked the ranking power I should have had.

3jUxO03.png

There's an example. You'd think disavowing tons of "juice" would make you lose rankings, and that was true, but now with links applying a negative ranking factor again, you can actually see an improvement, as shown above where tons of posts immediately jumped to #1 (dark green) and #2-3 (light green).

The good news is the disavow tool is working real-time and working fast and working good.

3) You have zero authority and trust flowing through your pages
  • Web 2.0's
  • Profiles
  • Citations
  • PBN's
  • Image Links
  • Etc.
All of those, minus the PBN's, and the others you mentioned can all be basically considered "profiles." Web 2.0's generally offer no juice (which is why you juiced them, understandably). All of these are likely nofollow, and if not then they're probably classified to send less page rank or aren't indexed at all. Some might even be noindex in the meta tags or in robots.txt.

The other unrealistic aspect to these profile style links is for them to be linked to contextually from a ton of other sites. It rarely happens in reality. And you did it with another PBN that's likely cheap, and that usually means sites are dropping out and being added constantly. So you've probably lost most of those links and the ones that are still sticking are likely telling Google to disregard the profiles altogether.

The PBN's you mention, whether direct or tiered, have probably "been made." Meaning Google knows what they are and let them exist so they can let spammers screw themselves over.

In my opinion, a lot of these "easy win" links, the profiles, web 2.0's, citations, etc... they're fantastic to have as long as you don't poison them. Getting them indexed is tricky, but that may not even be necessary so much as just getting them crawled. They can help you with your anchor text ratios and can be very legit links. Businesses and sites set up profiles all over the web. But they don't spam them with contextual links from PBN's. The best way to get them indexed and juiced is to interact on the sites so that internal dofollow links are pointing back to them.

The other thing is that these types of links, even when they exist in isolation en masse, don't offer authority and trust. Google shut that down a million years ago during the XRumer days. Just like they shut down social bookmarks, etc. These kind of "platform" links can easily be classified and devalued on the web before you even land the link.

But when combined with real marketing and real links, they can offer a lot. But their potential is zero until proven that they and your own site are both real versus splogs.

My suggestion:

Disavow all of the trash and any of the good stuff that is backed by trash. That may damn near be everything.

Then I'd find forums in the niche and create profiles with your link in them, and post on each one about 10 times in new threads and stickies if possible. Once that's done, return to each forum a month later and add your site to the signature once nobody is watching, with a URL anchor. I'd do this on as many decent forums as you can find directly related to the business.

I'd blog comment on business sites and sites directly related to the business.

I'd go out of my way to create one super-linkable asset, and get links and social signals to it from places like Reddit, Voat, Facebook, Twitter, etc... and have a link back to the most important page to rank.

I'd stop trying to rank the homepage for anything that's not brand related. Google prefers inner pages for non-brand terms. Yeah, you can pull it off and I'm doing it right now on a site of mine but it's ridiculously a lot harder, even with an EMD.

Most importantly, link-wise, I'd start outreach by the boatloads to land guest posts on real sites or at least a mention of your super-linkable asset.

This project seems recoverable to me, but it's going to require a lot of footwork to get back to where you were. If you're funded decently, I'd talk to Steve (who posted above me) about his service.
 
@Steve Brownlie Thank for the reply man. The owner is fully aware of the off-site strategy, I agree that guest posting is a way-way better solution than paid post on PBNs, but you know most of the times the budget make the rules of the game. As far as it comes to the T1 PBN links, I'm trying to get the highest quality possible. Cheap ones are on T2. I guess I will start to look for guest posts. As you mentioned, most likely G doesn't like the current links.

I know that you are deep in the guest posting game, so I am curious about if you try to push bloggers to post anchors by your choice or you let them pick what they think is most suitable?

@Ryuzaki, thanks for the helpful post. I was aware of this random document patent, and I knew it happened few times during the process. I acted like it was nothing. It's true that this is not the best way to rank a site, I've ranked sites in the past doing similar strats, now obviously I should change my approach radically. About the contextual PBN links on T2, I agree that this is highly unnatural, but this used to work on many occasion. I am tying to move away from this model, it is a bit hard when I work with limited budget, but it's time to adapt.

About the rankings on the home page, all of the competitor's sites target the primary KWs there, even thought everyone has separate services pages. I guess in this niche Google prefers to send the users straight to the homepage.

I will start disavowing some links, I'll do it in batches and see how it goes. As you recommended, I'll get involved in local forums and start doing outreach and hopefully will bring the site back to light.
 
I know that you are deep in the guest posting game, so I am curious about if you try to push bloggers to post anchors by your choice or you let them pick what they think is most suitable?

For my clients, naturally, we try to do whatever they ask, but my recommendation is honestly to keep things as natural as possible and not try to influence the anchors much at all with the sites who choose to link to you. In some competitive, national, search categories I totally get why people want to be more aggressive, but in local I've never found the need to 'push the envelope' very much at all so I tend to opt to recommend as close to 'full safety' as you can go.

One less thing to worry about going wrong in future if you haven't over optimized your anchors!
 
Back