One of the more common questions you see asked in forums is “How many links do I need?” in order to rank for a certain keyword phrase. The question doesn’t really have a specific numerical answer, rather the number of links you need is proportional to the number of pages competing for the phrase. Less competition requires fewer links to rank. More competition requires more links.
Consider a made up word like “grabandle.” None of the major search engines are returning any results for the word and with good reason. I’m probably the first person to type the made up word “grabandle” on a web page. I’ve now typed it twice and would expect that not too long after this post is indexed it will rank #1 for the word.
Assuming no one else decides to adopt my new made up word and use it, this post should require exactly zero links to rank for the phrase simply because the competition for the phrase is also zero. This post should soon be the only relevant web page for the query “grabandle.” No competition requires fewer, and in this case zero links to rank.
Now consider the phrase “real estate.” Google is estimating there are 549,000,000 results for the phrase. Looking at the #1 result for the phrase Yahoo shows 352,000 links into the site and 249,000 into links to the specific page ranking.
I won’t expect this page to rank for “real estate” any time soon and neither should you. More competition requires more links.
Competition vs. Links
If we were to graph how much link juice or linking power you need given the competition that graph might look something like:
Starting at the origin with zero competition you need zero link juice. As the competition increases so too does the link juice required to rank. While I’ve made no attempt to actually plot how many links (and the associated linking power) are needed for various phrases, I would think the link juice necessary would increase exponentially as shown on the graph. The curve, of course, might look different in reality. The main point to take is that as competition increases the link juice needed to rank well also increases.
Please note the use of link juice or power in the graph as opposed to the number of links. Links aren’t equal. Links have different signs of quality. A link from a forum signature is clearly less valuable than a link from the New York Times. There isn’t a 1:1 correlation between links, which is why the graph above refers to link juice instead of the shear number of links.
Strategies for Ranking
The above is all fine and good, but probably nothing earth shattering. Common sense alone should tell you it will take more links to rank when there are more sites and pages wanting to rank for a given phrase. So how can we use the information?
If you’re starting a new site in a competitive market such as real estate you should realize it’s going to take a considerable amount of time to rank for the more generic phrases on your topic. It’s going to take a long time to build the links required to rank for the most generic “real estate.” Knowing that you should understand that a long term strategy is going to be necessary to rank for that phrase. The tactics of the moment are not going to be your salvation. You’d sooner want to look at what will be working tomorrow or next year.
If your business model doesn’t allow for such long term results you’d need to look at other ways to market your site such as pay per click advertising. You’d also look for long tail keywords you could rank for much quicker.
On the other hand say you do have a site that’s acquired a good deal of link juice over the years. Even if you’re not already in a specific market you may be able to rank well for phrases simply by using them on the page. You already have the juice, or near the juice, to rank. If CNN covers a story there’s a good chance the story will rank for the phrases used in the article, based on all the links that site has.
Odds are your somewhere in between the extremes. You have some link juice, though not enough to rank for anything you want. You have enough to rank for some phrases, but not quite enough to rank for others. How can you tell what you can and can’t rank for?
Server Logs Reveal Link Juice
Your server logs will tell you what you’re ranking for. They’ll also tell you what you could be ranking for. As an example this post on centering with css routinely picks up search traffic for long tail phrases that include words like “css,” “center,” “centering,” etc. A phrase like “css center,” while not the most searched for phrase is likely a phrase the post could rank well for with a little optimization for the page. Not necessarily a great phrase, but one that has a little more competition and can bring a few more visitors each month.
My search logs tell me that I have enough, or close to enough, link juice to rank for the “css center” phrase should I want and some on-page seo might well be enough to bring in search traffic for the phrase.
Your real estate site might be ranking well for “San Diego, real estate for sale” and other similar phrases. That might also indicate you’re ready to rank for “California real estate” or something something somewhat larger with presumably more competition than San Diego real estate.
By showing what you’re ranking for now, your server logs are intimating what you can rank for next. Your server logs will tell you qualitatively what you can do with your current link juice.
Getting back to the original question, “How many links do you need?” the question again isn’t one that’s going to be answered by a specific number of links. You don’t need a specific answer to the question, though. Consider the competitiveness of the phrase and look through your server logs to determine if you potentially have the juice to rank for the phrase.
If you do, then start optimizing for the phrase. If you don’t look for phrases a little further along the search tail. Choose a phrase that is closer to the head than you can currently rank for, but close enough that you can rank for with a little work. Once you’re ranking for those somewhat more competitive phrases start the process again.
Do you use your server logs to tell you what you can rank for?
If you liked this post, consider buying my book Design Fundamentals