Key tips – Web Analytics for blogs

I’m frequently asked about “which metrics should I use to measure my blog?” and my answer is always the same “It is not possible measure all the blogs with the same information structure, it really depends on the blog profile”. I answer that inviting people (you) to take action on their own project/s. However if you are still interesting in that information, here we go:

I hope you find useful the following tips, but please don’t forget to consider your blog profile and peculiarities before designing our own information system.

I recommend you to think your website as a pipeline where things flow through it:

1- As bigger the beginning of the pipeline as much quantity of “something” you can put in there.

2-As longer the pipeline as larger the quantity of “things” that can remain inside for a larger time of period.

3- And last but no list, if the end of the pipeline is broader means that much things can flow out, letting another thinks come inside the beginning of the pipeline.

So we will split the information system in three stages.

1-    Inbound traffic

2-    Interaction.

3-    Outbound traffic

The idea is to understand how the traffic is flowing through your system (blog), identify where the weakest link is (restricted capacity resource) and how generating any action on the weakest link could modify the performance of the rest of the blog.
So let’s take a look to some metrics you could you use to measure each of the above mentioned stages:

1-    Inbound traffic: In this stage you may need to define which sources are driving more traffic to your site with the lowest bounce rate, so let’s take a look to some interesting metrics:

a.    Bounce rate per referring source: This metric is simple, just take the bounce rate form each of the referring sources. It will tells you if some of the channels are not driving qualified traffic to your blog. It is important to consider that if your blog presents all the content in the home page most of the traffic will be bounced, but that doesn’t mean that your blog is not working.  In that case I suggest you using the following metric (I prefer this one but the following one is a good backup):

b.    Average time on homepage per referring source: To make a “time per page” metric useful I suggest using a benchmark. It could be yours or from another similar blog. If you want to use a benchmark from your site, you can take an average from the last 6 months. This is a comparative metric, which means that is helpful just if you compare the time per visit from one source to another, if you do so could replace the previous one very good.

c.    Average visits per referring site: This metric allows you understand if your blog is being properly linked, which means a balance between quantity of sites and visits from each of those sites. Having a huge amount of traffic coming from just one referring site (just to give a drastic example) represent a huge risk to your site. First, because if something happen to that other site your traffic is done, and second, because you wouldn’t have enough relevant links to get a properly organic position.

d.    Quarterly referring sources trend: Calculating the last 3 month trend from each referring source is mainly useful to determine if you are achieving the expected results. On the other hand this metric will also help you understand what you would expect in the future (medium and large term) from your site. If your web analytics tool doesn’t provide information about percent variation here you have the formula

Percent variation = ((Visits n) – (Visits n-1)) / Visits (n-1)

Please ask me if you need more detail on this.

e.    Online visit rate: This metric (daily visits/daily rss subscriptions) will tell you if your blog is getting more online visited or rss visited. If the second, then forget about pageviews ;-) . What’s good? whatever your objective is. Information allows your to measure things that are important to your site, but what is good or bad is determined just by you.

To be continued…on Thursday I promise ;-)

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.