Organic Search SEO

SEO traffic predictions – how to calculate keyword value

Thursday, December 27th, 2012 by Steve Masters tagged ,

SEO crystal ballWhen we talk to clients about SEO, they often ask one (or both) of these two questions.

  • How long will it take for me to get X,000 visitors per month?
  • What ranks can you promise me in three months, six months, 12 months etc?

The obvious answer is to ask the client to tell us how long is a piece of string. There is no definitive and no right or wrong response, but we know why the clients ask the questions. They are trying to get some understanding of the process in a bid to put a return on investment (ROI) calculation on their spend. The piece of string does vary for each client, but there are some things we can do to predict timings and traffic.

Work out how fast you can get higher rankings

Start with keyword research

You need to start with keyword research. It is the equivalent of deciding which property to rent for a real world business. Keyword volume is like footfall. Keyword research is about finding the search phrases that your target customers use the most when they are looking for a business or a product like yours. It is also about understanding intent – do search phrases signal an intention to purchase or a seeking of knowledge? Using the free Google Adwords Keyword Tool (or any keyword tool you prefer), you can discover likely search volumes for specific phrases.

In addition, look at the level of competition for those phrases. The lower the difference between search volume and competition, the easier it will be for you to rank for that phrase.

Sort out your on-page optimisation

Once you know the most valuable phrases to focus on, you need to make sure your site is written and structured to reflect those phrases. For example, do you have a product category called ‘fashion items’ when most of your customers actually search for ‘designer clothes’? The usage of your top search phrases around your site, in conjunction with a sensible hierarchy, is an important factor in your SEO success.

A crude way to predict the level of traffic you can expect

This is a really grey area but it is a good way to give you an indication of traffic. If your business is in selling financial services and you are hoping to double your website traffic, this exercise will show you whether that is possible using the keywords your site is built around. Sometimes you can achieve good rankings but total demand is just very low.

First, some industry statistics

Using anecdotal evidence, based on research findings published online, I make the following assumptions:

Search user behaviour

In addition, there is evidence of the numbers of clicks for each position on page 1 of search results.

Search results page 1 and 2 - where the clicks goYou might see different numbers published over time but this is the breakdown I am using for my calculations.

How to calculate traffic estimations

Now take the keyword research you have done, with the volume of searches expected for each keyword per month and the level of competition. Create a table with the top 10 positions. In each position, calculate the expected traffic as follows:

Search volume x 62% x rank%

So, you are using only 62% of the search volume for page one, then working out how many of that 62% each rank might be expected to turn into clicks.

If you now consider the level of competitiveness for a keyword, you will be able to think about how long it will take you to reach a certain position. This is why SEO experts now talk about a wider range of long-tail keywords being more lucrative because you can get faster and more reliable results.

Does it work?

In my tests across keywords for a few sites, using existing data, I saw a few keywords drove more traffic than expected. In most cases, the actual traffic was between 40% and 50% lower than the results predict. If, therefore, you want to use this model to predict your likely traffic for a range of keywords, I would knock off another 50% to give you a more realistic business expectation.

Also, seasonality comes into play. On one site I tested, in the month I was testing, the website was in its best sales cycle, so demand was a lot higher. In another case, demand was way below average. So, to get a really true picture of your traffic, you could go fiurther and break up your data by expected volume for each month of the year. My model uses Google’s annualised averages and is intended as a guide.

Where it works is if your boss, or your client, is expecting greater performances. This model allows you to translate level of demand into likely traffic figures.

Share this article

About the author

Steve Masters

Steve Masters

Steve is Services Director for Vertical Leap. He started professional life as a magazine journalist, working on music magazines and women's titles before becoming a web editor in 1997, then joining MSN to work purely in online publishing. Since 1999 he has worked for and consulted to a broad range of businesses about their digital marketing. Follow on Google Plus and Twitter