Take the 2-minute tour ×
Webmasters Stack Exchange is a question and answer site for pro webmasters. It's 100% free, no registration required.

I'm new with SEO. I want to start with on-page optimization as it is quick and easy to implement.

I will start with HTML structures such as <title>, meta, headers, etc. and then set up robots.txt and sitemap. However, I will not touch up on content as we have other people (writers) to do so. Before I begin though, I want to have a benchmark of the old site so that I can measure how effective my on-page optimization is.

So what do you measure on-page optimization with and how?

share|improve this question
    
You could get a seoscan before and after. But the good ones will cost you something. If you have budget, sitebeam.com works pretty satisfactory –  Martijn Apr 16 at 10:13

3 Answers 3

Unfortunately, you most probably won't be capable of measuring the SEO performance of the page before and after on-page optimization. The reason is you can't freeze the search engines for a period of time for your keywords. Search engines continue to calculate positions during the time where robots (web crawlers) index your page page for the first time and index your page after your modifications.

You can try to compare positions of your page for your keywords before and after but you have no guarantee that positions changed because of your modifications.

share|improve this answer
    
Thanks for pointing that out Zistoloen. Yes I realize now that perhaps there's no clear cut way to measure as page ranking changes constantly. So would it be safe to assume that on-page optimization would increase page ranking but by how much, cannot be predicted? –  alds Apr 25 at 11:10
    
I would say you can assume that on-page optimization would increase a page ranking as long as 1. you respect guidelines of search engines regarding SEO and 2. you think about your visitors when modifying. –  Zistoloen Apr 25 at 11:20

When I want to do SEO changes and measure the results, I change half the pages on the site. This works best if you have a site with lots of similar pages that get similar amounts of search engine traffic. To make it work:

  • Divide your pages into two groups, randomly if possible. I generally use the content id for each page and do even and odd.
  • Exclude pages that typically get a lot more SEO traffic than other pages (for example the home page).
  • Make the changes on one group of pages but not the other.
  • Let the experiment run for a while (at least two weeks, I usually wait a month).
  • Measure the amount of search traffic to each group compared to a previous time period and see which group changed the most.
  • Be careful of outliers. For example, if you have a page that went viral or got press coverage during the experiment, you probably want to exclude it from the results. Whichever group that it is in will likely get a boost that has little to do with the changes you made.

Here are some types of changes that you might make using this methodology:

  • Title templates: Test "product at brand" vs "brand - product"
  • Templates for meta description, H1 or other areas that are likely to alter search engine rankings
  • Page layouts
  • Keyword density

To measure the results, use the following metrics:

  • Number of search engine referrals -- Most important metric
  • click through rate (CTR) -- available from Google Webmaster Tools
  • Bounce rate - to ensure that the changes are not turning away users

Pro tips:

  • You can have more than two test groups. Just divide the page id by something larger than two.
  • Dividing the page ID by prime numbers gives you multiple test groups that can be run simultaneously. For example, divide by 7 and put mod 0, 3, 6 in group A; mod 1, 4 in group B; and mod 2, 5 in group C. The test groups won't be of equal size, but that can be accounted for by looking at percent improvement of the metrics rather than the absolute value of the metrics.
share|improve this answer
    
Thanks Stephen. Nice tip on how to select pages to test (divide into two, odd/even, remove outliers). But testing this way is not limited to on-page optimization right? This test strategy is also done when updating content correct? Also as an SEO newbie, I may be mistaken that the effect of each on-page tweak can be measured. But as you and @Zistoloen pointed out, effects can only be observed, not measured, and done so over time. –  alds Apr 25 at 11:05
    
I've updated the answer with additional information about which types of changes this will work for and what metrics you would want to measure. –  Stephen Ostermiller Apr 25 at 11:54

Our SEO firm uses Moz's On-page grader to measure it, but that tool mostly makes sure you are doing basic things right like making sure the keyword you are optimizing for is used throughout the page. Your best bet is probably to use a tool like Majestic SEO to check what your off-page optimization is like (links, domain reputation, etc.) and then take the change in that into account when you look at your before and after rankings. That should help you isolate the effect of your on-page efforts.

share|improve this answer
    
Thanks for the input Mike. I wish we are a profit site so my company would invest into these SEO tools. I'll look at Majestic SEO's free plan and perhaps learn other points to consider. –  alds Apr 25 at 11:17

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.