Some of the KPIs in Search Engine Optimization are plain and clear: most people, for example, measure the ranking positions for a few keywords. Or they look at the overall traffic. Or they analyze the number of links. But can you improve on measuring those KPIs? I believe so, and in this article I will present some pitfalls in measuring only some KPIs, as opposed to having a global view.
- # Number of links from individual pages – This KPI can be easily manipulated by having just one or two web sites with a huge number of pages link back to your site from footer or sidebar elements. Once you do this, you will get a huge number of pages linking back to you. But Google knows how to evaluate this very well, so he will not take into account the vast majority of those links.
- # Number of linking domains – Generally, this KPI is a very solid one. Attention, though, as it may still be manipulated. For example, if you hire an SEO contractor to add links to you, that contractor can add your web site to a 100 of business directories, a 100 forums and 100 social networks. You will see 300 new domains linking back to you, but there are a few problems: first of all, the majority of those links will come from domains with low authority. Then, quite a few of them will likely have the “nofollow” directive appended to them. Finally, social network and forum spamming is considered a SPAM signal by Google and you risk having an automated or manual penalty. Just to make myself understood – this KPI is, overall, a good one, but you still need to make sure that the new domains you get links from are quality ones and do you good, as opposed to creating problems.
- # Number of words in an article – Sometimes in the past, this technique used to work. If you wrote a very short article (50-100 words, for example), you would generally get poorer results than by writing a longer article (500-1000 words). But search engines got more and more sophisticated and now, as far as I can tell, they generally look at indirect signals telling them about the quality of the article itself (bounce rate, time on site, clickthrough rate, links pointing to article, social sharing), and also on signals which confirm the quality of the article (topic keywords present in the article, level of comprehension of the text, spelling errors, things like that).
- # Ranking positions – Generally, this is a very good KPI to consider. But you also need to take into account things like – how many keywords do you actively measure? (it is generally better to measure more keywords, to get a more balanced overview of your ranking status) Also, have a look at organic traffic (traffic derived from SEO activities) – does it rise, at the same time with the new keywords? Then, you should look into topical keywords, so instead of focusing on a short set of keywords, try to see if your pages rank high for terms which are related to your top pages.
- # Number of pageviews or # Number of sessions – when analyzing this, especially for a web site with small traffic, you should be filtering your own (and your client, if this is the case) IP Addresses. You should only take into account traffic from other sources than your own. Another thing to look into is the type of traffic. Instead of just looking at the overall traffic, you should segment your traffic into traffic from organic (Google) searches, paid traffic, referral traffic, traffic from social networks. If in one month you have a large budget for paid traffic and in another one you don’t, this will affect your overall traffic. Also to be noted is the fact that one type of traffic indirectly affects visits for another one. So, paid traffic affects organic traffic in Google Analytics to some extent. Another thing about this KPI is the quality of traffic. You should focus not only on getting lots of traffic, but getting a good quality traffic. Yet another thing
- # Number of conversions for ecommerce web sites – this is quite an easy one, you should also monitor the average value of an order.
- # Cost per conversion – when you monitor this statistic, you should always compare it to # Average lifetime value of a customer. These two should be in balance.
- % Percentage of returning visitors – when you look at this statistic, you should be segmenting your traffic. If, in one month, you put a lot of emphasis on getting new customers via paid advertising, and you ignore campaigns for your long term audience, your statistics will be influenced by this trend. You should be comparing traffic from organic searches, paid search, referral traffic, social networks, etc.
- Using a single KPI – whatever your SEO campaign is, my advice is to have a look at more than a single KPI. Even if you’re just interested in profit, and you do only SEO campaigns, nothing else, you should still consider evaluating other criteria. For example, you might ignore # Number of newsletter subscribers, which will bring you clients on the long run. Or you might do good SEO on a temporary basis, but on the long run this might not help you, because you only focused on short term (example – running only social media campaigns, contests for social media, things like that).
Also see: Performance Magazine | What happens when you measure the wrong KPIs in SEO?