Measuring Your Keywords

/

As a search analyst, I’m often obsessed concerned with how many visits a website has received from search engine traffic. It’s one way to measure the success of an SEO project. Even better, I like to measure how many visits came from non-branded keywords (because those are the people who don’t already know you – new blood!).

But there’s one thing I want need to know and track on an ongoing basis that GA just doesn’t do. I need to see how many keywords send traffic and whether the number of keywords (NOK) is increasing or decreasing. Fortunately, GA does at least tell us how many keywords sent visits for a specified period, it just doesn’t let us see how that number changes over time.

Why are number of keywords important?

There are a couple reasons why I need to know whether the NOK sending traffic to a site is increasing or decreasing over time. Generally, it helps to measure progress during an SEO project. It’s one more indicator of success, especially if a main piece of the project involves creating content.

Indexation

If the NOK is increasing, that’s a good indication that the number of pages of the site being indexed and/or ranking highly is increasing. If you’re creating a lot of new content (blog posts, video, recipes, articles, etc.) it’s good to see if your efforts are paying off.

Of course, other factors can affect your NOK. Maybe you’ve changed your navigation menu or site architecture, or maybe you’ve been doing a good job of getting links to deep pages within your site. The effects of those endeavors will likely be seen in your NOK.

Early Alert

Monitoring your NOK also gives you a heads-up on site issues. If you see your NOK suddenly decrease, you can quickly look into any site changes that were made that may have caused a drop in rankings or indexation. No changes? Maybe your site’s been hacked (check Google Webmaster Tools for your site and Fetch as Googlebot).

Variety

Is your blog like a broken record? Do you keep talking about the same thing over and over again? Check your NOK – is it stagnant? Maybe you need to talk about something different, or at least from a different perspective. By introducing more variety to the content on your site or blog, you’ll be able to target different keywords. Even if your blog or site is very narrowly focused, I’m sure you can find some aspect of your niche that you haven’t discussed before.

How to Monitor Your NOK

Like I said before, GA doesn’t make it easy to monitor your NOK over time. To do that, you’ll have to do some manual labor in Excel. It’s nothing fancy – no macros or formulas involved – just some simple data entry and charts and graphs. Here’s a sample chart of what you can get, followed by the step-by-step instructions:

number of keywords that sent visits to a website over time

1. Open Excel and create a new spreadsheet. In the first column, label the first row “Date” and the second row “NOK” (or Number of Keywords, if that’s easier to remember).

2. Determine how often you want to monitor your NOK. I recommend at least once a month, but if you are actively engaged in link building or if you’re making some significant site changes, you may want to check this more often. Of course, you can go back in your analytics data as far back as you have it, so in addition to keeping track of it moving forward, you can back fill previous months. At this point, you can fill in the specific dates in the columns of your spreadsheet.

3. Find in the numbers. The only thing left to fill in on our spreadsheet is the number of keywords that sent traffic for each month. This is easy enough to find in Google Analytics. Go to the Traffic Sources > Keywords report. Set the date range to the month (or week or day) you’re interested in, and you’ll find the NOK right beneath the graph, like below:

(click to enlarge)

4. Fill in the numbers. Take the NOK from that report and enter it in your spreadsheet. Then wash, rinse and repeat. In a few minutes, your spreadsheet should look similar to this:

5. Create a chart. You’re almost done! Now all you have to do is create a chart based on that data, which is easy enough to do in Excel. From the “Insert” tab, click on Line chart:

6. Select your chart data by clicking on Select Data. In the pop up menu, for “chart data range” just click on the cell with the first month’s number (1,578 in the example below) and drag to the last month’s number (or you can hold “shift” and click on the last number).

7. Fill in the details. You can pretty-up the chart by editing the labels for the horizontal axis and series name, and in no time, you have the chart that I showed you at the beginning.

Now, if someone could create a Firefox plugin or web app using the GA API that would do this automatically, I’d be a happy man. Any takers?

Do you measure keyword traffic? Tell us about it in the comments below!

Jim Gianoglio is a Senior Analytics Consultant. He works with implementation, analysis and training of Google Analytics and Google Tag Manager. Before focusing on analytics, he led the SEO campaigns of Fortune 500 companies in the insurance, retail and CPG industries. Things you didn’t know about Jim: he’s biked from Pittsburgh to Washington DC in 41 hours, roasts coffee beans and has done voiceovers for TV commercials.

  • http://www.joelpinkham.com Joel

    Great post – I’ve only just started measuring number of keywords but it made me ask more questions.

    I now track how many of each length by setting up a couple of Advanced Segments to include only keywords that match regular expressions
    1 word: ^bw+b$
    2 words: ^bw+b bw+b$
    3 words: ^bw+b bw+b bw+b$
    4 words: ^bw+b bw+b bw+b bw+b$

  • http://www.joelpinkham.com Joel

    And just to add to that – I also track number of keywords / landing pages from organic search.

    This gives me an indication if my increased number of keywords is coming from more pages ranking, or more volume coming into the existing number of entrance pages.

  • http://twitter.com/jgianoglio Jim

    @Joel – not one, but two excellent comments! Thanks for adding to the conversation.

  • Mehdi

    Thanks Jim.

    I already made such an analysis and the visit trend was the same as the NOK trend. So I wonder if it is worth it.

    What are your obeservations about that?

  • House

    Overall this is a good measurement. However, GA results get skewed with publishers who do SEM. There is a Paid/Non paid filter but it is known to be buggy and often your keyword count will include SEM keywords. Alternatively, you can also track this in Google Webmaster Tools which I believe is more accurate.

  • http://twitter.com/jgianoglio Jim

    @Mehdi – I would pay attention to both. Your visits trend and NOK trend may not always be the same, and if/when they diverge, that will tell you something about your site. For instance, if you see that you NOK is no longer increasing but your number of visits continues to increase, that would tell you that either 1) more people are clicking on your listing in the search results (probably due to higher rankings) or 2) that your rankings have stayed the same, but the search volume for the keywords you rank for is increasing (possibly due to seasonal trends).

    @House – The only “bugginess” I can think of with regards to the paid/non-paid filter in the keywords report is if the ad campaigns aren’t being properly tagged (especially if they’re running ads on Bing/Yahoo). Certainly, not tagging your search ad campaigns will skew the numbers here.

    I’m a big fan of Google Webmaster Tools too (just gave a 2 hour webinar on them earlier this week). But I would be never use those numbers for this purpose. They just aren’t accurate, and you never know when Google will change the tool or remove that feature.

  • http://analyticsedge.nextanalytics.com Mike

    Hey Jim, I automated the report for you (http://analyticsedge.nextanalytics.com/2010/08/monthly-metrics-measuring-your-keywords/) and I included a few more metrics as well, but Mehdi may have a point — none of the 5 sites I tried showed any valuable insight from the analysis.
    Also, tracking just the number of keywords on your chart without showing the number of visits may lead you to a wrong assumption.

  • http://twitter.com/jgianoglio Jim

    @Mike –

    I can’t wait to try your report generator (I have to sign up for a Next Analytics trial account first) – this could definitely save some time!

    You’re absolutely right – the number of keywords should never be looked at in isolation. I should have done a better job explaining that in the post (lucky for me our astute readers fill in the holes).

    I generally look at 5 metrics on a monthly basis (with additional metrics based on the client and specifics of the project):

    1) Total visits to site (from all traffic sources)
    2) Total visits from organic search traffic
    3) Visits from non-branded organic search traffic
    4) Number of keywords
    5) Number of referring sites (excluding “mail” to filter out visits from web-based email)

    As for whether or not there are valuable insights from this information, I can think of a few cases to shed some light.

    Let’s say you’re adding articles to your site in an effort to pull in some additional long tail search traffic. Over the course of a month you add 100 articles. Then you look at your number of keywords sending traffic and see that they are increasing (along with non-branded search traffic). That indicates some success – you’re increasing your search traffic by adding new content which is pulling in visits from keywords that you previously didn’t get traffic from. If you only looked at just the number of visits from search, or even non-branded visits, that’s a good indication that those 100 articles are increasing your traffic, but it’s not as accurate. It could be that some other page on your site just started ranking well for a high volume keyword. If that were the case, then you’d see an increase in search traffic, but not an increase in number of keywords.

    That would certainly make me dig deeper into the issue – maybe the 100 articles are too deep in the site and not getting enough link juice. I would check to see if they’re being indexed, check my internal link structure, etc.

    Another case is if you have 10-20 head terms (high volume keywords) that make up about 30% of your organic search traffic. Things are going along fine until one day you notice that your non-branded search traffic has plunged. By looking at the number of keywords, you can get a sense of the cause of the plunge. For instance, if your number of keywords is relatively steady, then you’ve probably lost rankings for a number of your head terms. However, if you number of keywords has dropped, then you have to do some more digging. Maybe you’ve lost rankings across the board, or maybe just to the long tail.

    At the end of the day, I always feel better having more information than less.

    Thanks for the comment and many thanks for automating the report!

  • http://www.engage-digital.com hugh gage

    By coincidence I was looking at this just last night. I also think it’s useful for paid search campaigns where the campaign may have been broadened out and in doing so could bring in some less relevant terms that could suppress conversion rate while still pushing up conversion volume and a lower CPC (depending on the terms of course).

    One thing I did notice when I looked at both Google paid search and Google organic search for July year on year…. the total NOK count shown in the bottom right of the report was higher than the combined totals when I looked at each month individually i.e. not using the data comparison for July YoY. Do you have any thoughts on why that might be, am I missing something obvious?

    Cheers.

  • http://www.ethanandjamie.com Ethan Gardner

    @Joel – Thanks for the regex snippets. It really helped me get started. I’ve modified yours a bit so I can get a report of any search traffic that used a keyword with 3 words or more.

    The regex that I used was:
    ^bw+b( bw+b){1,} bw+b$

  • http://www.universityusability.wordpress.com John Wedderburn

    Thanks for this post. Maybe it would be interesting to look at particular segments of traffic (for example, visitors from specific geographical regions) and see if there are differing trends in their keyword growth – just a thought!

  • http://twitter.com/jgianoglio Jim

    @John –

    I like your thinking! Using analytics without segmentation is like baking chocolate chip cookies without the chocolate chips. Or butter. Or eggs.

  • http://www.universityusability.wordpress.com John Wedderburn

    @Jim – exactly. The more I think about this, the more interesting it gets – you could combine your changes in keywords with traffic which converts, this would probably get rid of a lot of noise and, hopefully, show the results of any SEO work etc.

Contact Us.

LunaMetrics

24 S. 18th Street, Suite 100,
Pittsburgh, PA 15203

Follow Us

1.877.220.LUNA

1.412.381.5500

getinfo@lunametrics.com

Questions?
We'll get back to you
in ONE business day.