>> Home > Mach5 Resources, Support, and Idea Bank Home | My Mach5 | | About Us | Purchasing  
  
Idea Bank Home

  

Explore M5 Subscriber
Web-based list management made simple!

Simple subscribe tool
Opt-in confirmed
Track reads and clicks
Custom matching look
Free for up to
        5000 contacts


Explore M5 Mailer
Take the bulk out of email! TM
Use your own data
Personal email merge
Handy email templates
Scriptable with
JavaScript, PHP, VBasic


Explore M5 Analyzer

Get great site statistics directly from your web server log files.

Optimize site design
Track campaigns
Performance Metrics

 

 

 

Analyzing Web Site Visitor Trends

So you've spent a good deal of time and money developing a slick web site for your business. You're selling high-quality products and investing in a Cost-per-Click ad campaign. Unfortunately, sales are disappointing and you suspect that you're overpaying for many of your visitors. What do you do? A web site analytics package can turn your questions, doubts, and suspicions into concrete, quantitative answers, and help turn visitors into customers.

This article provides a brief introduction to web site analytics in three quick-to-read sections:

  • Baseline visitor metrics. How many users visit your web site on an average day? Do these users come from links on other web sites, one the of the major search engines, or do they simply type your URL into their browser?
  • Planning the trip. Once you've established a few basic facts about your site's visitors, it's time to map out how they move through your web site. You probably have an idea of how you expect visitors to move through your web site, but do the data support your assumptions?
  • Cost-per-click searching. Paid search results can be a cost-effective way to attract visitors to your web site. How much is your average visitor worth? Are you paying too much for your visitors, or could you afford to up your advertising spending and attract more visitors?

If you are in charge of web site design, no doubt the answers to these questions are very important to you. That information resides in your web site server log files, and a good log file analysis will yeild a treasure-trove of important information about how you may improve your website and marketing efforts.

Baseline Web Site Metrics

Before delving into more advanced analytics, it's important to develop a basic picture of your website. How many visits does your web site receive on an average day? Where do these users come from? These statistics are a crucial first step in any site optimization.

Hits, Page Views, and Visitors

A 'hit' represents one request for a single file from your web site. A visit to a single page on your web site typically spawns many hits. When a user types your web site's URL into her web browser, the web browser issues separate requests to the web server for the HTML page, any images that might be embedded in that page, and any CSS files used by the page, among other things. Each request to the web server is known as a 'hit'. It is common to quote web site usage statistics in terms of hits per day, but because a hit is generated for each image, CSS file, etc., this is a poor measure of actual usage. Page views – hits directly to HTML pages – are a much better measure of actual site activity.

The definition of a hit or a page view is very concrete, but the concept of a 'Visitor' is much fuzzier. Web sites interested in tracking visitors typically assign each user a cookie that contains a unique identifier. The 'cookie' is stored on the user's computer but is included in every hit on the site by that user. Most web analytics packages will utilize this cookie information to determine an approximation of the number of unique people who visit a web site in a given day. However, your site may not place a visitor cookie. In addition, some users configure their browsers to not accept cookies, and cookies can sometimes get lost or garbled – determining the number of visitors is more prone to error than determining the number of hits or page views, but the best web analytics packages have developed reliable heuristics that give excellent approximations.

This graph, generated by FastStats Log Analyzer, is dominated by a weekly, cyclical pattern; this web site sells business-related products, and receives dramatically more hits during the weekday than during the weekend.

As you make improvements to your marketing and search word advertising campaigns, you should see a corresponding change in your visitor traffic. So it is important to get a good baseline of visitor traffic before making any changes.

Where are your visitors coming from?

Referrer information can provide a wealth of insight into your customers. Every time a user clicks on a link in an external web site, the address of this external web site is recorded as referrer information. This information – properly analyzed – can be mined for information on search engine optimization, general marketing, etc.

Important questions to ask are:

  • Which web sites drive the most traffic to my web site?
  • Which search engines drive the most traffic to my web site?
  • What percentage of users are from CPC search advertising? What percentage are from natural search engine flow?

Referrer information, unfortunately, isn't perfect. Many browers omit or mis-report the referrer information -- a significant percentage of all visitors to a web site are likely to have no referrer, even if they arrived at your web site by clicking a link in their browser. Thankfully, there are reliable workarounds for any limitations in the accuracy of the referrer data. One common trick is to have each advertising campaign land on a separate page on your web site. For example, your Google ads can land on http://www.mach5.com/products/analyzer/?from=google, your Yahoo/Overture campaigns can land on http://www.mach5.com/products/analyzer/?from=yahoo, etc.

Unfortunately, this technique only works for links that you have control over, such as cost-per-click advertising. Natural search flow -- visitors driven to your web site by Google's search engine -- may arrive via any page, not just your designated landing pages. You can only discern natural search engine flow by decoding the referrer field. Its important to understand the limtations of all of the different pieces of the pipeline -- your users web browsers, your web se

Planning the Trip: Where will your visitors go?

Once a visitor has found your web site, the next step is to convert them from a visitor into a revenue-generating customer. To do this, you need quality products, good sales material, and a logical web site layout. A web analytics tool can help you tune your web site layout and rework your marketing material in order to increase revenue.

You probably have a mental picture of how you anticipate visitors moving through your web site, but it's important to use your web server logs to put hard numbers on these hunches.

Planning the Trip: The Visitor Plan

To test to see if your site is working the way you think it is working, first formulate a visitor plan. Write down where you think you would like to see your visitors go. Then, take a look at your web site analysis to see if your visitors are, in fact, going where you need them to go.

Step 1: Determine your site entry pages. What site do you plan on having visitors land on?

Step 2: Plan the exit. Where do you want your visitors to actually end up? What's the goal? That page should be the end of your flowchart.

Step 3: Plan the road map. What pages do you see your visitors going to next? List them out. If progressing toward an ultimate goal, visitors may jump to the goal along the way. Draw out a flowchart of where you think your visitors are going to go. Put percentages on the pages in your flowchart to anticipate your user behavior.

Planning the trip: Analysis

Visitor Path Analysis

The above screen shot is a capture from a web analytics program, FastStats Log Analyzer. The screenshot graphically depicts how visitors click between pages on the web site. This web page needs tuning – well over half (58%) of all visitors leave the web site after viewing this page.

Site Entry Pages

A Site entry page is the first page that a visitor sees on your web site. It's the landing page, either from a search engine, a referring partner, or in rare cases, from the URL entry bar. These are some of the most important pages on your web site – they must be carefully tuned to retain the maximum number of customers and direct them to the information they are seeking.

Step 1: Determine start page. This page will be one page of many site entry pages, where a visit starts. If you run a basic web site analysis, you can pick out the site entry pages. If you want to drill down on entry pages from a particular search engine or search phrase, you can filter for that phrase and search engine first. You can then see if visitors are really landing on the page you think they are landing on.

Site Entry Pages

The above table, from FastStats Log Analyzer, shows the top entry pages for a web site. How effective are these pages at retaining customers? Let's examine the statistics and see:

Site Exit Pages

In this particular example, index.html, the #2 site entry page, is also a top site exit page! Of all hits to index.html, the majority (58%) result in a site exit. This page needs to be tuned to be more effective.

Referrers

Referrers

The sites that direct your web site traffic – referrers – are the life blood of a web site. Links that drive traffic to your web site should be cultivated.

Search Phrase Performance

Scenario Analysis in FastStats

To set up Scenario Analysis in FastStats, just edit your project and add a scenario.

A scenario consists of one or more page hit steps through your website. In advance, you should know what steps you would like to track, for example from home page to product page, to details page, download, and purchase.

In the Scenarios setup box:

  • Click on the Add button
  • Name your scenario
  • Add steps that match
      - /index.html
      - /productdirectory/index.html
      - /downloads/product.html
      - /purchases/complete.html
  • Run your project!

Scenario steps also support wildcards, so you can match whole classes of pages across directories, if your web pages are named appropriately. For example, if all of your widget landing pages have "widget.html" on the end of their path name, you can set up a step for "*widget.html"

If you are spending money on search placement, you definitely have an interest in knowing how well different search terms perform. A web analyzer can help you answer questions like:

  • How many new users do paid and unpaid searches deliver to my web site every day?
  • How much are these visitors worth? What percentage of users actually complete an order form and purchase a product?

The second question is crucial for pricing cost-per-click ad campaigns like Google's Adwords or Overture's Precision Match. One way to determine an approximation of the average value-per-click is to measure the percentage of users who enter from Google Adwords and eventually purchase your software.

Scenario Analysis

This screenshot, from FastStats Log Analyzer, shows the percentage of visitors from Google who ended up downloaded and then purchasing the software. 510 users entered the web site during this particular day. Of those users, 83 (5.49%) downloaded the product, and 3 (0.20%) purchased the product. If the product in question sells for $500, the equation to find out the value of each click is (510 clicks from Google) * (X cost per click) = (3 purchases from Google users) * ($500.00 per purchase). With these numbers, the average Google click is worth $2.94 per click.

Home   |   Site map   |   Analyzer   |   Mailer   |   PopMonger   |   Resources   |   Customers and Partners  |   Contact
Mach5 is a trademark of Mach5 Development, in use since 2000.