John Webster is a fellow I recently met through this blog. He came across some of my search engine optimization articles and instantly made contact with me. Turns out John is a super fan of web site marketing and SEO and runs his own firm in Australia called AIM IT.
John sent me through a solid report on analysing the key metrics in website traffic reports that he gave me permission to share with you. If you are at all into web site marketing I think you will find the materials in this report valuable.
Because there was so much content I’ve broken down the report into two parts. You can also find further details about John and his business at the end of the report in part two.
You would not dream of trying to run your business without a range of reports that show you how your sales and promotional activities are performing and your web site should be no different.
I find that many people are unaware of what web site traffic reports do and the essential information they provide so you can assess your online marketing performance. This article addresses the key issues.
Web site traffic reports are usually generated by programs that reside on the web server that hosts your web site. There are specialist services that offer alternative solutions but I am going to focus on the basics.
Every time someone goes to one of your web pages, their browser asks for a copy of the web page file and all the individual image files that are displayed on the visible web page. The request essentially includes the address of the requester’s computer and the web server responds by forwarding a copy of the file(s) to that specific location on the world-wide-web.
A web site traffic program simply sifts through and analyses all these individual requests and responses and provides the information in a number of easy to read report formats.
Every time you dial up for Internet access, your Internet access provider assigns your computer a temporary web address and it is this address that is used to identify your computer when your browser requests web page files.
When you close your web connection, the temporary address you had been allocated becomes available for use by another subscriber to the same access provider service, so it is not possible for your web site host server to track the identity of any individual visitor.
Tracking individual users can only be done if the visitor accepts “cookies” and registers for some service when they visit a site, but that is another issue I don’t intend to address here.
The term “hit” can cause confusion. If you say that there were so many “hits” on your web site, you are saying that this is the total of all files that were requested by all site visitors and in this use of the term, it includes all image files.
A web page file (also called an HTML file) is not like a Word document in which you can embed images. Web page files do not contain any images. All they contain is text – the visible text on your web page, the HTML code text that tells your browser how to display your web page and instructions to get a copy of any image files from the hosting web server that are to be displayed by your browser as an integral part of the visible web page.
In this context, “hits” has no relevance to web site performance as the number includes all the image files as well as the web page files accessed by site visitors. To double your total site “hits”, all you need to do is double the number of images on every site page.
However, some web site traffic reports may use “hits” in a context that is relevant. Some may report visits to web page files (also called URLs or unique resource locators) as “hits”. In this situation, because all other file types are excluded, “hits” are relevant as the reported number is showing how many times a web page file was requested.
The best use of your reports is for trend analysis.
Web site traffic reports are 100% accurate but they are not 100% real. Confused? Following are some of the factors that cloud the reality picture.
Visits to your web site are not all real people. There are many programs released on the web to visit web sites. These include the “goodies” like search engine (SE) robots that automatically crawl the web requesting copies of web page files and the “baddies”, like “sniffer” robots that may visit to seek email addresses for “spam” email generation. You would expect your traffic report program to be installed to exclude reporting visits from the major SE robots but it is not practical to identify and exclude the many other possible non-real visitors.
The visitors by country traffic report is another one that has limitations. Some Internet access providers can assign their subscribers with temporary computer identification that does not enable interpretation by the traffic analysis program. This may result in 15% – 30% of visits whose country of origin cannot be determined.
Then there are the large corporations who often use worldwide networks. In this situation, any employee in any location can be accessing the web via the country location of the head office and will be identified by your traffic reports as originating from that country.
I am focusing on the basics and I regard this as the absolutely essential web site traffic reporting information you need to manage your web site and online marketing.
Online Web Site Traffic reports are typically updated daily or weekly and you would expect to be able to access current month and a monthly history of web site activity for the past year.
You will want access to reports for each month that show:
The big picture – Number of web site visitors, number of pages visited and pages viewed per visit
The significance of these individual reports is discussed below.
You would want to track the number of web site visitors, the number of pages visited, the pages viewed per visit and the pages viewed per site size at least monthly.
The other element you need to record is the size of your web site, its growth and when content was added, as this will impact on the pages viewed per visit over time.
The need for tracking the total visits and pages/download files viewed/accessed should be obvious but the pages viewed per visit and the pages viewed per site size, warrants some explanation.
Example – Getting it right
The following web site traffic figures are taken from a site that targets Australian marketers and corporate PR staff.
Pages viewed per visit, provides an indirect assessment of the relevance of the site’s content to the visitors. If the SEs are referring a lot of traffic to a site, it is quite possible that many visitors are clicking through but finding the site content not relevant to their specific needs. In this case, you might find a low number of pages viewed per visit, as these visitors will hit their “back” buttons pretty quickly.
The figures for this site suggest the content is relevant to many viewers and the consistent increase in pages viewed per visit suggests the site is targeting the desired audience. However, this metric alone can mask problems that the pages viewed per site size may disclose.
This site consistently increases the number of published web pages and by this analysis there was a 5% decline in the pages viewed relative to the size of the web site in Dec 05.
There are a number of possible explanations for this drop. Typically, there is a total down turn in web usage by Australians in Dec/Jan and the December figure may simply reflect this.
Another explanation may give cause for concern.
Google implemented major changes in Oct 05. They changed how their SE ranks search results and they approximately doubled the number of web pages that they index. When SEs implement changes of this magnitude, pages from your site that previously ranked high to relevant search phrases can drop dramatically in the results ranking and cause a significant down turn in your site traffic.
If this was the reason for the decline, then web site remedial action may have been essential.
This is one of those reports where “hits” = pages viewed or download files accessed.
Its primary use is to identify what areas of your site are considered most relevant to site visitors. If people do not go to the pages that are important to you, it may indicate a problem with your site structure, descriptive content or SE targeting.