≡ Menu

SEO: Calculating a Keyword’s Value

Using Analytics Data to Drive SEO

Now that Google Analytics is available to anyone with a website for free, it’s possible to bring an entirely new level of data-driven analysis to ongoing SEO efforts. This guide assumes that you have a basic understanding of analytics concepts and a functional web analytics platform that tracks conversion data that can tie back to the original source ( one of the most basic features). It doesn’t have to be Google Analytics, but my examples will use their reports.

So many keywords, so few pages

Websites are structured like a pyramid, with the homepage as the most important page, and the pages that get the most internal links within the site ranked above those with fewer links. One of the biggest challenges for a SEO is to figure out which keywords to optimize a given page for. By now, everyone should be familiar with the concept of long tail keywords – these are more specific search terms that get far less search volume than the ‘big head’ keywords, but are easier to rank for, convert better, and are used later in the buying cycle. It would be a mistake to optimize an important category page for a long tail keyword, and optimize a less important page for a big head keyword. This guide will show you how you can avoid these mistakes by allowing the data to drive these decisions. The best part of this is that SEO is an ongoing cycle, and once you’ve made your first round of changes you can go back and see how successful you’ve been. It also includes data on any PPC activity that you might be doing, and you can use the same data to feed back into your bidding on paid keywords.

The right metric

The first step is identifying the metric we want to use. in this case, the best metrics will have these properties:

  • Available in default reports within most analytics systems (or easy to calculate).
  • It should reflect the conversion rate, which can vary widely from keyword to keyword.
  • It should also reflect the average order value (AOV), which also varies.

First metric: dollars per visit ($/visit)

With this measurement, we can finally reach the holy grail of SEO – the ability to know exactly how valuable a keyword is for a given landing page. Some keywords get much more traffic than others, and it’s important to have a measure that isn’t tainted by your existing traffic patterns, and allows for rarely used keywords to shine, proving they warrant more SEO effort. Simply looking at revenue or orders, for example, could lead you into a cycle where keywords that you’re already ranking for will be given too much priority.

Dollars per visit = conversion rate * average order value

Dollars = orders * dollars
Visit    visit  order

You can also simply divide revenue by visits, it all works out to the same math. Luckily, this metric is available in all Google analytics reports that provide source tracking – under the e-commerce tab. For example, you can go to content > overview > landing pages, click the e-commerce tab, and select ‘keyword’ as the secondary dimension. This will show you visits, revenue, transactions, and average value, conversion rate, and per visit value ($/visit) for each landing page and keyword. I found it works better to export it and turn it into a big pivot table when dealing with lots of data.

Bringing it home: Keyword Value in $/month
Google’s webmaster tools continues to get better. One feature they added in the last couple years is called ‘search queries’. this shows you the top organic search queries at your site ranks for, along with impressions, clicks, and click through rate (CTR). These are not exact numbers, and should only be treated as estimates. This gives us the final piece we need to find the potential or real value of a keyword in dollars per month (we should remember that CTR is highly dependant upon the position on the SERP).
Keyword value ($/month) = dollars per visit * search volume * clickthrough rate

Dollars = Dollars * Searches * Visits
Month    Visit  Month  Search

I’ve done a few keywords as an example:

Keyword

Conversion %

AOV

$/Visit

Searches

Organic CTR

Rank

Keyword Value

Canon digital camera

7.98%

120.46

$0.10

110000

20.00%

2.9

$2,114.80

company name

14.89%

129.11

$0.19

1300

23.00%

2

$57.48

LCD TV

0.96%

75.36

$0.01

823000

0.10%

9.3

$5.95

desktop

1.00%

79.07

$0.01

60500

2.00%

7.5

$9.57

canon power shot camera

2.30%

77.93

$0.02

5400

2.00%

3.9

$1.94

What I like most about this calculated metric is how it accounts for any of the variables that can and do vary greatly. Some keywords get a ton of search volume, but they generally have weak conversion rates. Keywords associated with larger AOVs will benefit as they should, and it even gives us a method to calculate the value of X number of hops up in rank (e.g. moving from 10th to 2nd).

It’s important to remember that higher volume keywords will usually convert at a lower rate, which means there is always a trade-off to be considered when deciding which keywords to go after on a given landing page.

Using the Data

Now that you have a few hundred or thousand keywords with an estimated potential value for your site, it’s time to assign keywords to landing pages. I recommend optimizing at least two or three pages for each important keyword, and using combinations to optimize for long tail keywords. Most sites have a hierarchy where the category pages for important than the product pages beneath it – use that structure to map your keywords. The most important pages in your site should be optimized for the keywords that have the most search volume and/or competition.

A Note About Variable Data

 Web analytics is an inexact science, and it always will be. People clear their cookies, shop from multiple computers, and use a wide range of browsers and security software, all of which can disturb our tracking and generate strange results. You will find examples where a long tail keyword may be responsible for way more in sales than you think is reasonable. This is a tricky situation, since it is very difficult to tell what’s actually going on. I have found that the best approach is to use as many data points as you can, so the major trends emerge and are easy to differentiate from weird blips. Use a year’s worth of conversion data, for example, and you will get more reliable results than a month’s worth. It will also remove seasonality effects that can vary greatly from month-to-month.

The following two tabs change content below.
I have been working with computers and web sites for 20+ years, and have enjoyed mastering many areas of technology. I have been building websites for about 15 years, and working with NetSuite for more than 10. I have worked with dozens of small and medium-sized companies in that time, helping them to understand and leverage the latest tools to grow their business. My business is all about helping you to maximize your business, and I prefer to establish long-term relationships with clients who are dedicated to embracing smart ways to optimize and expand their business.

1 comment… add one

  • It’s appropriate time to make some plans for the future and it is time to be happy. I’ve learn this publish and if I may I want to counsel you few attention-grabbing issues or tips. Perhaps you could write subsequent articles regarding this article. I want to read even more issues about it!

Leave a Comment