Skip to content

May 19, 2010


Link Building Normalization

by Anthony Verre

Link Normalization: Continuous Link Building

Links.  A few simple words in hyper-text. Something so simple should be easy to come by, should have superfluous resources and opportunities. The problem is they do.

The problem is many novice SEOs, and the ones who are just posturing as professionals, build backlinks in a sporadic, dump-truck kind of way.

Pile on a 500 links today, sit back for three months, and then pile on 500 more. Seems like that might be the right tactic; would anyone really believe a site can grab 500 “natural” links a month, consistently? Off of static, crusty content nonetheless? Not a chance.

That’d raise an army of red flags from the engines, signaling there is “link manipulation” at hand.

You Got Flagged by the Search Engines

You’ve Been Flagged

The problem with that tactic? You just got flagged. That kind of “hit-and-run” link building technique is viewed more suspiciously and manufactured than if the website really was getting 500 links a month.

Search engines, rather the people behind the algorithms, are much smarter than people give them credit. Three or four years ago, the hit-and-run technique might have provided a decent organic boost. The crawling wasn’t nearly as frequent and as dense as it today. So the layoff might have been in order, depending on the number of links one was acquiring.

Faster, Deeper Crawl Rates. Oh My.

Today, search spiders are crawling faster, deeper, and more often. Even to low Toolbar PageRank sites. Consider the MayDay update (possibly part of the Caffeine overhaul) , in conjunction with what Andrew Shotland found concerning accelerated crawl rates. Fresher, more relevant results for every query.

The solution is to build your links in a consistent manner. It’s why I tell my clients one simple mantra when the link building begins: “Once we start, we don’t stop.”  And that’s the bottom-line. No matter what we’re doing, no matter the content we need to build, we continually build links.

Your Link Graph Rise-Over-Run

Once you can accept and understand that concept, it’s all downhill from there. It doesn’t matter how quickly you build your links. It doesn’t matter how slowly you build your links. As long as it stays consistent, and you create a link graph rise-over-run that ends up looking like this:

Creating the Right Link Rise-Over-Run

Time for Some Link Research Data:

It’s not enough to just to tell you link building normalization matters. Google’s going to show you. I’ve taken some SERP slices from a highly competitive market: “Miami Real Estate”.

Searches were done with “personalization” off (i.e. not signed-in to my Google account, web history off, and &pws=0 tacked on to the end of the query.)

SERP Stats

For this test we are looking at the 2nd and 3rd position sites, what separates them (what distinguishes them) and then a result from the 4th SERP page, position 42.

Miami Real Estate SERP Stats

click to enlarge


  1. Site’s domain ages are around same level of oldness
  2. Site’s with the highest link domain are below the 2nd SERP result
  3. Link Balance for the 42nd result is quite disproportionate (567 links of the 51,765 to the homepage)
  4. “Miami Real Estate” in domain may be the reason it’s on top of the others. Also the oldest of all the domains.

One might draw from this that 2nd result must have more powerful, relevant links, and with the domain name incorporating a the keywords and it’s age, it’s understandable why this ranks better than the others. Yet, better to be thorough and make sure this is the case. Then you look under the hood.

What the Link Normalization Graph Looks Like

I wanted to see if the pace of link building for these sites has been steady throughout their histories. It’s not unusual to see link totals that high, but for real estate sites that old, it drew my attention.

SERP Results link building normalization chart

There you have it. Look at the link surge of the 42nd result. It’s that link-spike from July of 2009 to October of 2009 that doomed the site. Moreover, you can practically pinpoint the time when this site starting getting some “SEO help” (if that’s what you want to call it) by way of link building: July 2008.  The 2nd result has had a very stable rate link discovery for years, save the small surge  in early 2008. The 3rd result was as erratic as the 42nd result, but with the majority of the real shoddy link building occurring in 2007 into early 2008.

Doesn’t Matter Where Your Links Come From

Everyone of these sites has “junk” links. Stuff that’s non-topic and irrelevant. It may be time to bust a link building myth. Here’s a look at sample swatch from the second result.

click to enlarge

And there are more swatches of links like this throughout this site’s link catalog.  The same is true of both the 3rd and 42nd result. So we can effectively toss away this as a determining factor in the positioning. If all the link profiles are nearly the same: massive amounts of junk links with a few relevant, on-topic links, then we can zero this out.

The truth of the matter: it doesn’t matter where you get your links from (to a degree). If your domain age is old enough, and you’ve built enough goodwill, you can pretty much link anywhere you please and it shouldn’t negatively impact your position. But I caution you on this. The evidence is here to make that statement, but I certainly wouldn’t take my fresh site and pollute it with bullshit links. This sample has fairly old sites (4-5 years respectively), and that may be the leverage you need to link with impunity.


In my estimation, it is the rate of normalize link building, outside of the second result’s very targeted domain and age, that really made the difference in how these sites stack up in the SERPs against one another. This is not to say that the onsite efforts of each are necessarily better than one another, and were not taken into account (therefore nullified here), as this was a offsite analysis.

The data shows that normalized link building trends can influence the position of your site.

  1. May 19 2010

    Well written but this kind of rubs me the wrong way in light of how anecdotal the analysis is. I have no problem with the idea that constant vs spiked link growth can be a flag for the SE’s, but to make the assertion that this is was the causational factor behind the different rankings in this situation seemed weak. It seems like there should be a higher burden of scrutiny for making these kind of confident assertions. Bottom line, there are alot of variable at play and I didn’t think this was air tight enough. Not trying to rain on your parade, I just don’t want to the bar lowered any further than it already is in the SEO world for people making unfounded claims about how the SE’s operate.

    • May 19 2010

      No rain taken. Admittedly, onsite factors were not considered when evaluating the placement of these sites. That’s going to carry a lot of weight in determination of SERP placement (i.e. how good is the internal link structure, how well the site is being optimized, is the content as optimized as it could be, etc.) Was the link growth of 42nd result the sole reason it got pushed to the 4th page? Probably not. But, I would stake that this is a determining factor. And, let’s take into account, this is a single snap-shot in time.

      That being said, perhaps this is a momentary lapse for that result, pushed back because the “spaminess” factor tripled in very brief time (massive link growth in very short period of time, then dropping off to nothing). I don’t think the claim is unfounded and it’s definitely something to keep your eye and make sure it’s on the level. In this single case, I’m making the argument that between all the offsite factors, the “normalization” of the link building seems to be a deciding factor. All those sites had tons of non-topical, irrelevant links, an abundance of them. Some had “powerful” directories, and some did not. Again, I’m not saying this was the only factor, but IT WAS a factor in this case.

      Thanks for the great comment Jordan, look for to a response. Cheers.

  2. May 19 2010

    “Again, I’m not saying this was the only factor, but IT WAS a factor in this case.”

    Fair enough, that clears it up for me and I would definitely agree.

Trackbacks & Pingbacks

  1. Tweets that mention Link Building Normalization « The Milwaukee SEO --
  2. SearchCap: The Day In Search, May 19, 2010
  3. Weekly Search & Social News: 05/25/2010 | Search Engine Journal
  4. Link Building this Month (05.2010) |
  5. SMX Advanced 2010 Seattle Recap - What I learned |
  6. May Linkbuilding Roundup | Search Engine Optimisation, Web Design and Photography Blog

Comments are closed.

%d bloggers like this: