Servage Magazine

Information about YOUR hosting company – where we give you a clear picture of what we think and do!

Use statistics to improve your web results

Friday, October 23rd, 2015 by Servage

statisticsAwareness of your users’ needs is the starting point for any improvement. This information can be gathered from multiple sources, e.g. satisfaction surveys or questionnaires. Instead of being gathered specifically, it could also be collected during more natural interfaces with customer, e.g. during support requests. However, it is also very meaningful to gather this information systematically based on users’ behavior, and nothing is more clear than raw statistics data from users’ behaviour on your site.

You may already be collecting this type of information, but are you also using it? Really using it? Meaning that it is nice to look at visitor numbers and most popular sites on an analytics page, but what does it tell you? Are you using the information to improve anything?

Starting to collect statistics data from your website

If you are not already collecting statistics data from your website, then you should start doing so quickly. The fastest method is probably to choose an existing solution like Google Analytics, KissMetrics, Kiwik, Clicky, Woopra, Parsely or similar. Usually you swiftly validate your site ownership and then just need to paste in a few lines of code, whereafter your site is pushing statistics to the provider’s servers for review in a dashboard. Most providers are similar in features, but you should look into them before selecting, in case you have specific needs.

Visitors and page views

This metric is probably the one that web site owners look at first, and find very curious to follow, in hope of a constant rise. The metric is a pretty straightforward and direct indication of your site’s and content’s popularity, however it does little explaining of it. Therefore overall visitors and page view numbers are good to show you the current state, but you need to look elsewhere for reasons. Why do you suddenly receive more visitors? Why do you suddenly have less? It is like a stock chart. You can see the current value and historic development, but it is unlikely that you find meaningful explanations for the numbers by looking at the chart alone.

Content popularity

Looking at content popularity is a good way of understanding user behaviour. Do you have few blog posts that attract all the users? That is a common case, and you should try to generate more content similar to this (quality). However, if you have a more broad spread of content popularity (e.g. your content is all somewhat similarly popular) then focus on creating more bulk (quantity).

Search terms

If you realize drops in visitors for previously popular content, then you can investigate which search words or references used to lead users, and why they are no longer doing so. Example: If the search team “best content ever” used to have your site as first result, but now it is something else, then you know where to focus your energy: Get back on top of the result list for that search term.

Click streams

Some analytics tools allow you to track users’ click streams. Study them to identify pages in the process that have larger than average dropout rates. Example: If many potential new customers cancel the order process on step 3, then you may reconsider the content and design of that particular step.

Split testing

Whenever you need to understand or improve a given scenario, then use split testing where you expose some users to one variation, and other users to a different variation of the content. Thereafter you can evaluate and choose the most efficient variation. This method also solves internal discussions about what Call-to-action elements are better. You can simply test it and find out for sure.

Use statistics to improve your web results, 4.8 out of 5 based on 4 ratings
Categories: Business

Keywords: , ,

You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

No comments yet (leave a comment)

You are welcome to initiate a conversation about this blog entry.

Leave a comment

You must be logged in to post a comment.