The term ‘link spikes’ has been around a long time, usually to describe unnatural link patterns that Google uses to put the smack down on overly aggressive link builders.

Using the term in this way is correct and yes, creating crazy link spikes in your link profile will put you in trouble.

What people don’t realise is that it isn’t just positive link spikes that can cause issues but also negative ones. Spikes from lost links can be as damaging as spikes from link growth.

Of course losing links can mean lost rankings, however in my experience lost links can also trip filters and mean a specific page on your site will just never rank, no matter what you do.

Negative Link Spike Case Study

I had the unfortunate opportunity of experiencing a filter on an affiliate site of mine. After renting a few links for 6 months I decided to pull them down, knowing I had plenty in reserve to keep my rankings.

These links were site wide and all had the same anchor text, when they were removed I lost 52,000 links from 3 domains. Technically this should have been a good thing as the links were completely unnatural and obviously paid.

Two days after the removal of these links my site dropped rankings, not just for the target keyword but for any other search terms containg the target term. Elsewhere organic traffic was growing, and growing well, but this keyword and variations weren’t even in the top 100, you can clearly see the effect on traffic below for the page where the links were pointing.

The first thing I did was try and replace the lost links, I didn’t think this would work but wanted to see what happened, the answer? Rankings dropped even further. Now I had created 2 unnatural link spikes in a row and new my rankings would never recover.

Sometimes when you begin working on a new site you may come across ugly site wide links that you think need cleaning up, however do this with extreme caution, negative link spikes are as dangerous as positive ones.

First of all let me confess the term ‘fresh rank’ has been stolen from fellow SEO blogger Justin Briggs, I am going to refer to one of his excellent posts throughout the rest of this one.

You will no doubt know about Google’s new QDF upgrade, an algorithm tweak designed to get you to ‘fresh’ content quicker, rather than bringing up old static results.

You can see an example of it here;

They’re not site links but links to fresh content on the BBC for the search term ‘football’.

Google has stated that this affects around 35% of search queries, don’t get that mixed up with searches. Now that is all well and good but from my point of view I want to know a few key points;

1) How does Google decide what is fresh?

2) Is the link graph involved when deciding ‘freshness’?

3) How do links from these ‘fresh’ pages influence rankings for the taget website?

I wrote a really short post a few months ago based on fresh links vs text links vs links placed in old content. The results clarified that links in fresh content had a more significant impact on rankings.

However, I feel the need to delve more into this, as I think this strategy is one of the most important link building tasks you can undertake and will help you cement long term core rankings.

How Does Google Determine Freshness?

Justin wrote a great post on this on his blog and if you really want to delve into this you should definitely go take a read, I just want to touch on some of his points and then try to understand how we can use it for link building purposes.

Document Discovery

I think it is safe to assume that the discovery of a document through Googles crawl for the first time is enough to indicate freshness. There is a little bit of debate around this, but nothing that affects the take aways too much, does Google count it as fresh when it is first crawled, first linked to, first mentioned in a social capacity, first indexed…? No one really knows and the truth is it is probably a mixture of all those factors.

Proportion of Change

Is a document fresh if it hasn’t been discovered before, or is it if a document has significantly changed since the last crawl? Do Google give the content a score on a sliding scale?

This is important to know, if the freshness of a document determines the power of a link from that document then we need to know what this scale is, Google is never going to give anything away, I doubt a new FreshRank toolbar is going to emerge, however we can test this through our link building efforts.

Fresh Rank

This brings us nicely onto ‘fresh rank’ a term I am fully attributing to Justin. In the paper Systems and Methods for Determining Document Freshness, it describes a method of passing a freshness score between pages.

So just like PageRank is passed between pages so is FreshRank, so whether or not your landing page is ‘fresh’ will depend not only on the changing content on the page but also by the freshness of pages linking in, hence we have some kind of freshness score.

Is Freshness More Influential Than PageRank

I use the term PageRank very loosely and only to describe the authority of a domain or page based on the quality and quantity of links it has pointing to it.

We all know how PageRank travels around the web and through our sites, its long been the currency of the web, the more links a site or pages a site has pointing to it, the more value a link from it will pass.

Now before I go into this any further, let me first declare this has not been tested or researched in any way, my opinion is based on working with an SEO department that manually builds over 5,000 unique links every month.

The PageRank Model

Value is passed from page to page based on popularity.

FreshRank Model

I am a firm believer that the above model is already in place and has been for sometime and that Google use a combination of FreshRank and PageRank to determine the ranking of a given page.

No one knows the exact calculation but it could be that a link from a fresher PR1 page is worth as much as a static PR6, maybe that is overselling it a bit but certainly getting multiple fresh links every month can be just as effective as acquiring 1 high PR link a month. I guess the ideal is to combine the two.

I’ve worked with many sites over the years and the best way to impact rankings is by creating a fresh link profile. The ultimate combination would be creating a fresh link from a high PageRank (authority) domain, combining high PageRank with a FreshRank strategy, is a sure way of dominating your industry SERP.

Diminishing Value

The one problem with a ‘fresh’ link building strategy is that it is likely to diminish in value over time, therefore efforts have to be ongoing.

PageRank flowing through a link is going to be more consistent than Fresh Rank flow, a document rarely consistently acquires a large quantity of links naturally, however once a document has a certain amount of PageRank it generally keeps it, that is of course as long as the links remain live and and the PageRank isn’t pushed through multiple 301′s.

As you can see from the above model links are carrying PageRank and FreshRank, however what about 6 months later;

Overtime the freshness of the linking documents will diminish, no doubt everytime Google crawls them it will give them a new score. If only we could build lots of links that are from fresh, high domain authority pages that are going to be continually linked to for the rest of their existance

Overall Thoughts

I have seriously thought about researching this in more depth, but from the results we see on a daily basis it is pretty much a given. SEO strategies that involve the consistent development of links from fresh pages will almost always achieve higher rankings than those that don’t.

Only Google will know exactly how this works and one day they may be willing to shed a little more light on it but that won’t stop me and more SEO’s implementing it.

Would be great to hear some debate on this or examples of tests/research.

OK, I’m going to cover some basics here but time and time again I get asked about monitoring indexation levels and what factors affect them.

There are quite a few misconceptions in this area and a lot of people rely on the wrong set of data, they also rely on old techniques to improve this, so let’s have a go at discussing some of the main areas.

HTML & XML Sitemaps

What does a sitemap really do? My opinion? Absolutely nothing in terms of getting a site indexed. It’s an auditing tool to monitor and test the architecture of your website, it doesn’t matter how many sitemaps you have, if the structure of your site is poor you’re going to have low indexation levels.

This is especially true of XML sitemaps; however what about HTML site maps?

Again in my opinion HTML site maps are over rated and used to cover up poor navigation, or used in the hope that they will create a magical ‘thumbs up’ signal to Google. The truth is most HTML site maps are not user friendly and consist of a pile of links, spread out over many pages, especially if you have a large site. With Panda hitting websites hard the last thing I would want on my site is a load of pages filled with html links tipping me over the low quality threshold.

The only time HTML site maps are effective in my opinion is if they are genuinely helping users navigate their way through a site and should include descriptive text as well as links to the various areas of the site, however if you own a website with 100,000 product pages the last thing you want to do is create a site linking to them all. The way your site is structured should help users and Google find all those products easily and effectively.

So to round up, yes use an xml sitemap to audit your site; don’t create a HTML site map unless it is user friendly and in place to help those users find their way to important parts of your site.

Try using multiple XML sitemaps to monitor the performance of different areas of your site.

Performing a Site Search

We’ve all done it and still probably do it, however this is a really inaccurate way of monitoring how many pages are actually indexed on your website. You can do one search and a search a minute later that fetches a different result.

However, if you know there are roughly 500,000 pages on your site, then doing a quick search can give you a very broad understanding of how well you are being indexed.

So go easy using the site: operator

Using Analytics

This is the best way of understanding not only how well your website is indexed but also the quality of those pages, it’s been spoken of many times before but let’s go over it again.

Login to google analytics, go into traffic sources and select google / organic, then select a secondary dimension of ‘landing page’, this will show you how many pages Google sent traffic to over a certain period of time, monitoring this figure every month gives you a really good indication of the indexation levels of your website, I would use and monitor this figure over anything else.

So there we have a few ways of monitoring but how do we get more of our site indexed?

Flat Architecture

Sorry to go back to basics but want to cover it for a sense of completeness.

First thing, don’t get confused with getting your site crawled and getting it indexed; these are two completely separate things. Once Google finds out your site exists I have no doubt it will crawl your whole site at some point, however in my experience getting good amounts of your site indexed comes down to one factor and that is trust.

Having a flat architecture is all about creating the shortest route possible to the pages on your site, how many clicks are your major pages away from the top level?

(image from SEOmoz)

It’s simple really, the closer a page is to the top level the more trusted it is going to be and therefore has more chance of being indexed.

It all comes down to links

It’s easy to get a page ranked by building lots of anchor text rich links to a page, however if you have done this before you will also have realised it has very little impact on the overall trust of a website.

I speak with clients about this a lot, we point out an overall increase in traffic and indexation levels and the client will say ‘but you we’re trying to rank us for X not the other keywords that drove traffic, this was just natural growth that would have occurred anyway’.

Trusted, quality links will affect the overall organic traffic to a website and are absolutely essential for seeing a continual growth in organic traffic.

Yes by all means build optimised links to your site and go after rankings but you have to incorporate a strategy for getting links from the best sites in your industry, the most trusted sites, without them your website will never perform to its full potential.

The above are the main factors but site speed and quality of content since Panda are having more and more of an effect on the indexation of a website. Check the speed and make any necessary changes to rectify speed issues, also check for duplicate or pages with low quality content.

Remember Panda is a ‘threshold’ based algorithm, removing as many low quality pages from your site as possible will could be all it takes to sort out Panda issues.

As ever would love to hear your thoughts on this in the comments.

If you run an ecommerce site you will have no doubt been frustrated with Google in the past due to the small amount of data it’s been able to provide concerning your conversion funnel. Many will have turned to coremetrics or omniture, however Google have pulled their finger out and changed their attribution model and are giving much more data.

Google multi-channel funnels is a game changer in my opinion, I just want to flick through a few features but you can read more about it here and here

If you have ecommerce set up you will need to navigate to the new analytics view and click in the ‘my conversions’ drop down, select multi funnels and you’re on your way to analytics heaven.

Which Channel is assisting??

Most analytics systems credit conversions based on the first or last click within a session.

First click = PPC > Organic > PPC > Referral > Conversion (PPC Conversion)

Last click = Organic > PPC > Referral > Organic > Conversion (Organic Conversion)

So that’s how it usually works but what if you had the ability to see every player in this chain, every channel that was assisting in pushing conversions to your site? This is what Multi Channel Funnels does;

This pretty chart shows which channels assisted more often.

You can also break this down into source/medium, the chart above shows Google organic produces 8390 conversions from the last click, however it assists in another 6027! Just using the last click is seriously undervaluing the traffic.

But hey what about the first click, how did the visitor first find the site?

Again Google organic has 8849 last click conversions but 9724 on the first click!

Popular Conversion Paths

Again this is broken down nicely:

You can also see time lag between first interaction and conversion as well as path length in terms of clicks.

It’s all absolutely priceless and in terms of us SEO’s whether in house or in an agency it is a massive help in terms of proving the power of Google organic traffic.

Every ecommerce site I have looked at so far shows Google organic traffic is anywhere between 75% – 125% undervalued!

Happy days Seriously you need to read about this.

After seeing the traffic drop off on a number of my sites and discussing the Panda effect this morning with Patrick, I decided to do a little more research into the U.S. drop off and how it is hitting other websites.

We looked at over 20 different sites and some of the Ecommerce sites ‘stuck out like a sore thumb’!

This is one of the websites in question and it was a very similar picture across multiple UK based e-commerce websites, up to a 50% drop in US based Google organic traffic as of the 24th February.

As of yet UK traffic has not been affected, however Panda hasn’t hit the UK yet and once it does this picture is going to be even more devastating.

Can You Do Anything About It

Well if you have lot’s of unique content you have nothing to worry about, however E-commerce websites could be hit hard as they have multiple pages that have content on them that is duplicated all across the web, this is what is causing the above.

The above should act as a warning to all UK based E-commerce sites, if your US Google traffic looks similar to this over the last month then get ready for wave two to hit:-

- Switch any spare resources to re writing content that you believe to be duplicated
- Write 300 + words of 100% unique copy for any products you have
- Invest in encouraging more user generated content

Patrick has published some data on Blogstorm and there was a nice post on SEOmoz talking about the positive impact UGC has, especially post Panda. Panda is expected in the UK any time now, creating unique, useful content has to be a priority for all sites moving forwards.

After doing a bit of traffic analysis across my sites this week it has become evident that my ‘Google’ organic traffic is down around 40% on certain domains, luckily these domains aren’t relied upon financially but the results are a little worrying.

There are no obvious trends in terms of keywords, however I noticed some big differences when segmenting traffic by location, here is a screen shot from one of my sites showing Google traffic from the U.S;

Google traffic from the U.S. is down 50% compared to traffic levels in February!

If I had been scraping content on these sites or auto generating it in anyway I could understand such a dramatic drop in search traffic. The fact is the 100 + pages of content on the site were written by me and are 100% unique in terms of structure and tone of voice.

So on the face of it this site should not have been affected by the update, however there could be a reason for this….

I am not protective of the content on the site and there is zero social interaction on the domain i.e. nothing has been shared on any of the main social sites. I have no doubt that the content has been duplicated all around the web and scraped left, right and centre.

I am pretty convinced this is the panda update and if I am right it means any genuine content driven site, that publishes 100% unique content, could also be affected by this update. How does Google understand what the master copy of a piece of content is?

Food for thought I’d say.

After all the shouting last year about the state of the SERP’s it seems, as predicted, Google have started this year with a bang, updates and fluctuations in the SERP’s are becoming much more frequent and maintaining top rankings for any sustained period is becoming more difficult.

I personally track around 25 competitive terms and the top 20 sites that rank for them, when I couple this with the data we hold at Branded3 it shows quite clearly how often Google are making changes to these terms.

If you look closely it is almost a weekly occurrence, apart from the odd website that controls position 1, those that rank in positions 2 – 10 are switching around weekly.

Now Google could be simply cleaning up a lot of the profiles they failed to take care of at the end of last year, however my feelings are that Google is finding an counting links a lot quicker than it ever has, that combined with caffeine is causing these massive fluctuations.

My only reservation is that we saw these kind of fluctuations last summer, Google was making a massive update and when it was finished the SERP’s became a little more stable. If this kind of ‘Google Craziness’ is here to stay it can only be good thing for SEO’s and clients as seeing the benefit of links should come around a lot quicker.

Faster Results?

If you speak to SEO’s in general they will tell you that you need to wait 3 months to fully appreciate the effects of link building and solid on page recommendations. Recently I have been noticing changes coming about a lot quicker, Google reacting to fresh links in a matter of weeks, not months.

Now I am not going to be bold enough to say anyone paying for SEO should see results in 2 weeks, because I don’t believe that, however once link building has commenced, generally speaking you should be seeing results a lot quicker. Of course if you already rank in the 10 gaining the top position is still going to be a drawn out process, this is simply down to the amount of competition on the first page of Google, however generally speaking, every link you build should make it’s mark a lot faster than it would have done 12 months ago.

If you’re anything like me you will have been in a position hundreds of times, where you a trying to rank for a given keyword and no matter what you do, you cannot break into that top 5 or hit that first position.

Part of you says ‘just keep going & build more links’, however the truth more often than not is that you simply aren’t building the right type of links.

I thought it would be a good idea to run through how I  like to analyse link data and establish where I need to build links.

So, I want to incorporate a little open site explorer, excel, pivot tables and vlookup. I’ll try and make it as straight forward as possible for anyone that isn’t that familiar with excel.

Collecting The Links

For the purposes of this example I have chosen 2 sites that rank in the top 5 for mobile phones, I am not going to reveal the sites as I can’t be bothered dealing with complaints, businesses spend hundreds of thousands on SEO but they don’t want anyone to know

Anyway head over to Open Site Explorer:

Filter out nofollowed and internal links then down load your links into excel where we can begin to play around with them.

Organising The Links

OK now we have all the data together there are only a couple of columns I want to use for the purposes of this example, Unique Domains and Domain Authority.

I prefer to use domain authority rather than page authority because more and more often I find this to be the most influential factor.

So…. to get the unique domains you will have to perform a ‘text to columns’ on the URL column, copy and paste it into another sheet otherwise things will get messy.

Once you have a long list of domains, input them back into the main sheet next to the url column and use the data & remove duplicates option to have a 100% unique list of links.

Add the Pivot

So you have the columns you need, select all and insert a pivot table.

Add domains into the values area (ensure it is set to count) and domain authority into the row labels area, this will display the data like so;

Simply add a chart if you want to visualise what you have put together which is literally a domain authority scale and the number of domains for each authority figure.

Once you have done all this for your site, you now need to perform exactly the same for the competitor ranking where you want to be.

Putting it together

So you should have 2 pivot tables, yours and your competitor’s, open a new sheet and add numbers 1 – 100 in column A (this is going to be where you combine the two tables).

In a separate sheet you need to add the two pivot tables side by side (same work sheet as the above structure), however simply copy and paste the figures not the 2 tables as we are going to perform a vlookup to put it all together.

Now in the new 1-100 table you have created you need to use vlookup to transfer the data for each site across, see the below vlookup query I performed, I don’t want to go through vlookup here, Richard Baxter is does a far better job than I could ever do so read his here.

Right, you have performed the vlookup and have your two sites side by side, you can now play around with the data and begin to understand what types of domains you are missing from your profile.

Or you could spice it up a little using an area chart

Well not really spiced up but you get my point…

Surely Just Building Links from Top Quality Domains Will Eventually Put Me At The Top????

Don’t fall into this trap, if you just go after the big fish all the time you will lose for 2 reasons…

A – They are harder to get so you will never get the quantity required

B – Since when did a link profile with PR5 links only look natural??

Be smart and build your links from the right places.

I just wanted to thank everyone for their support on the Best UK Online Marketing Blog Poll, I have no idea how I got to the top but I am flattered at the amount of support. If you haven’t had chance to vote yet I would really appreciate it

Thanks to all!

Many of you will remember the article I released on brand anchor text and how I thought it was playing a role in rankings, plus I insinuated that over optimised link profiles were been filtered to give a more natural look to the SERPs?

Well it seems I was wrong and anchor text is still the king after all, this shows the difference between correlation and causation perfectly and I am sure there are going to be many of you out there that are shouting ‘told you so’.

I apologise for misleading any one but as you will see Google have been playing tricks and I think they have a few more to play.

The up date that changed my mind on anchor text was played out in the UK around Halloween, Google’s idea of ‘trick or treat’.

But I Still Wasn’t Wrong

The research I performed was correct, the sites at the top of the SERP’s definitely had a better brand anchor text ratio and the sites struggling definitely had an over focused anchor text profile, however what I got wrong was the fact it wasn’t the ‘clean brand’ effect causing them to rank.

What Was Causing Them To Rank Then??

This is my theory, not to be academically scrutinised (to all you SEO scientists out there), it is based on what I have seen happening in the SERPs and I have a couple of examples to back it up.

Just to clarify, these are examples and in no way are meant to act as an outing!! I am just trying to prove a point and am not trying to get anyone penalised. If I don’t include them I will just get called out as having no evidence, so it’s a lose lose.

Anyway this is what I think has happened……….

A couple of months ago the sites at the top were home pages or top level domains, they were ranking with little or no anchor text and not even that many links compared to their heavily optimised competitors.

However those under optimised links had nothing to do with where they ranked, that brand focused anchor text was not Google giving them a pat on the back, there was something going on under the hood.

Behind the scenes these websites had internal pages optimised for keywords like ‘car insurcance’ , ‘no win no fee’ , ‘perfume’ , ‘loans’ amd many more. These internal pages were loaded with anchor text links, thousands of them (most paid) and when I say anchor text I meant the exact anchor text.

For whatever reason Google decided the home pages of these sites were more relevant to show in the SERP’s, but gave the home pages the ‘anchor power’ of the deeper page. Build highly optimised links to the deep page and watch the home page rank???

Now Google have switched to show deeper pages and the true ranking factor is there for all to see – massive amounts of anchor text from paid links.

Want Some Examples?

OK here we go……

Example 1

In the no win no fee industry the National Accident Helpline were ranking with their home page for the keyword ‘no win no fee’. It looked like they had no links and zero anchor text, however when Google lifted the veil and showed their highly optimised internal page, exact match anchor heavy and clearly paid, we began to see that they were ranking their based on a strong anchor optimised internal page.

Example 2

Now we have go compare, ranking at the top of Google for ‘car insurance’ a couple of months ago with what seemed to be a very under optimised brand focused home page, however again they had a deep page loaded with paid links and masses of exact anchor text pointing to a deep ‘car insurance’ page.

Again Google lifted the veil and we see the true reason why they were ranking so highly in the first place!! Mental..

Put Down Your Anchor Text!

Before you go rushing off building a gazillion anchor text links into that optimised internal page you have, first ask yourself why Google have done this?

Why have they begun to rank internal pages that have masses of exact match anchor text from clearly paid links?? I feel an up date coming???

I just can’t imagine Google letting link profiles like this stay at the top very long.

May Be I Am Wrong, Again

Of course I could be completely wrong and Google have simply decided to show the page that is most relevant rather than giving overall power to the home page. Maybe we should all start targeting specific internal pages for the keywords we want to rank for, maybe not.

This is your decision as a webmaster/SEO/online marketer, what are you going to do??

Does anchor text continue to dominate?

or

Does Google have a Christmas present for link buyers everywhere???

Any comments, discussions or answers would be appreciated

Google have recently updated the way they show link data in webmaster tools.

Historically the link data they have shown has literally been a snap shot of links to any particular website, the information was pretty much useless to an SEO at any level.

However, Google have now decided to break down that link data in a logical way and early today the screen shot of your links will have been broken down into unique domains.

This is a little scary as I looked down some of my linking domains with 1 giving over 30000 thousand links, maybe this is Google’s way of saying ‘watch those site wides people’

The most intriguing thing about this change is when you look a little deeper at those site wide links you have, Google is even pointing out how many pages of your site this domain points to. High volumes of links to the exact same page easily highlight manipulation techniques and makes it easy for Google to devalue these links.

Are these links from the same domain, pointing to only one page of your site being devalued? Is this Google’s way of telling us to cut down on site wide links?

It seems that WMT is becoming more and more insightful, with click through data, anchor text distribution and performance stats.

All these small improvements leave me wondering where Google are going with this, they have the data and ability to provide more SEO related information than any of the other SEO tools out there, does this mean the end of seo tool suites like SEOmoz??

As it stands the information is not quite accurate enough to replace any of the top tools, but maybe ‘watch this space’ is an appropriate phrase to use.

Just a short post but your thoughts are appreciated.