Can you juggle with web metrics?

Teaser Webmetrics

“Since I’ve got all these things I can measure, I’m paralyzed by all the opportunities.” (CEO of a service company) What do you know about your homepage? Did you ever think about defining web metrics?


Internet business is steadily getting more important. The success and profitability of a firm’s online presence can only be assured if it meets or even exceeds their visitors’ expectations. Web metrics can help firms in understanding, managing and improving their web systems. This article wants to show several possibilities to benchmark the quality of a web site. Sure, it will be interesting to have a first glance. Furthermore it shows up the expectable future benefits after investing in the improvement of a company’s online presence. Responsibles should really consider intensifying their online business. It’s worth it.


Authors: Felicitas Bauer, David Humbert, Lisa Dürr

Discover new paths

dv1611020.jpg

NetGenesis, specialised in providing web services for small to medium sized businesses, made researches about web metrics and they found out, that web systems produce a great number of data about user activity. The web managers they spoke with did not have the definitions, methods or means to use it effectively. NetGenesis also found out, that the managers are in need for common measurements. They uttered their frustration with not having the ability to get access at the information they know is sitting in their customer behavior data files, and not being able to link that information with the available wealth of off-line information. But they are also convinced about positive future developments. Each manager has clear perceptions what could be learned from the information they already had and might collect if only they had a few more resources and a little more support from upper management.

So lets speak about performance indices and what they can tell us.


On the picture you can easily see the hierarchy of web site activity:metrics.gifClick to enlarge


  • User: a customer (person) uses your Web site. He or she becomes a unique user.
  • Visit: each time this user explores your site, you receive a visit from that user. Thus a user may have many unique visits to your site over time.
  • page view: each visit by a user is composed of a series of pages that he or she reviews, thus each visit is composed of a time-ordered series of page views, otherwise known as click stream or click path.
  • hit: each page on your site contains many constituent objects such as body text, images, and video files. Each of these results is a hit to your web site, so each page view comprises many hits. That also means, that hits are an inaccurate measure of websites popularity. (source: NetGenesis)

papier.jpg

Another often used term is log file. It is a file that contains an automatically produced protocol of some or all of the information mentioned above, thus every request made to the server. With log file analysis tools, it’s possible to get a good idea of where visitors are coming from, how often they return, and how they navigate through the site. But calculating visits and users is more subtle and complex than counting page views. As a result a large company without a standard definition for visits or users is likely to be using many different techniques for arriving at these measures. When the results are brought together for comparative purposes, it becomes especially difficult to tell what is really going on. Are the differences reflective of actual customer behavior or are they due primarily to different measurement methods? (see also article: E-metrics, Business Metrics For The New Economy, published by NetGenesis Corp. in year 2000)

Visualizing web metrics


The graph below is an extract of such a log file analysis. It shows, how many visitors were on a site (users), the number of visits and page impressions.Excerpt of BA Mosbach web metrics backendtraffic.gifClick to enlarge


You can also find analysis tools in the web, for example webalizer and webtrends. It is interesting to have all these numbers, but what can one conclude out of it? NetGenesis also presents in its article some total new e-metrics, which are designed to provide insight into e-customer behavior and provide readily actionable information is presented: These tools are very helpful to interprete the data you have collected with log files.

  • Stickiness
  • Slipperiness
  • Focus
  • Velocity
  • Seducible moments

stickyness.jpg Do you stick your users?

The figure stickiness is related to duration and frequency of the user’s visit. Stickiness is a composite measure that gathers the efficiency of your content in terms of consistently holding users’ attention and allowing them to quickly complete their online tasks. In general, sticky sites are considered more effective than sites that are not very sticky. Stickiness is calculated by dividing the average number of pages visited in a given section through the total number of pages in the section. (source: NetGenesis)

slipper.jpg When your users should better slipper…

Slipperiness is the exact antonym of Stickiness. Some areas of your site are significantly better if they are very low sticky. One area a user should slip over is for example the Customer Support section for starters. NetGenesis found out, that “the last thing your management wants to see is how your customers have to look at lots and lots of pages and spend huge amounts of time reading about product fixes and return policies. In this case, you want the Customer Support section be slippery…” The meaning of slippery is, that the users should come in and should find quickly what they are looking for, and then go out. Analysis showed, that every additional click required to make a purchase represents another opportunity for the prospect to change his mind and back out. (source: NetGenesis)

focus.jpg Focus on…

Focus is another concept related to the behavior a user visits a site within a section. Suppose, there are 15 pages in a section. You can distinguish between those visits, where the users touch 2 or 3 of them – this would be called focused- and less or unfocused visits, where 8 or 10 sites are touched. If the average of users touch 3 of 15 pages you have a users focus of 0.2. Smaller values point out a more narrow focus, higher values point out a wider focus. What is better: narrow or wide focus? It depends. Narrow focus is good at a customer service area of a site, but perhaps not at an online auction section of a site. (source: NetGenesis)

speed.jpg Speed

How quickly a user moves from one stage of the customers life cycle to the next. The customer life cycle is a term used to describe the progression of steps a customer goes through when considering, purchasing, using, and maintaining loyalty to a product or service. Marketing analysts Jim Sterne and Matt Cutler have developed a matrix that breaks the customer life cycle into five distinct steps: reach, acquisition, conversion, retention, and loyalty. In layman’s terms, this means getting a potential customer’s attention, teaching them what you have to offer, turning them into a paying customer, and then keeping them as a loyal customer. Clocking the prospect qualification process gives you the average amount of time it takes for a member of a given market segment to pass from awareness to decision. With this knowledge the web designer can test alternative navigation techniques. (source: NetGenesis)

seduction.jpg Sweet seduction

Seducible moments. Who does not know the feeling of giving in a seduction? In front of a shelf with sweets, in front of a hyper super snip… In the internet, this feeling has also be evocated. It may be the rapid purchase button next to a desired product or an up-sell offer at the moment a customer is deciding between two service choices. The right encouragement or the right graphic might just do the trick. One should avoid so called rough patches, which diminish the momentum or speed at which a browser becomes a buyer. Rough patches often appear at those places where the shape of the customer life cycle funnel suddenly gets narrower because many users are abandoning the process. (source: NetGenesis)

buyers.jpg How many visitors become buyers?

Have you ever thought about how many visitors of your company’s web page really decde to buy one of your products after watching your range of products? If not, you probably should as that can show you whether further investment in your web business can increase your sales figures on the whole. The most significant benchmark for your economic efficiency in the web is the so called coversion rate. This figure is often used by internet marketers for reviewing the success of a web site. It tells them the number of visitors that finally generate a transaction. Typical values usually range between 1 to 5 %. That means that 95 to 99 of 100 the visitors leave your page without doing a purchase, filling in of an online application form or simply ordering further information on products or services.

Of course there are some superior sites which are able to impress with much higher levels of conversion rates. Guess what online shopping portals come along with conversion rates up to 10 %? To most people there comes amazon.com to their mind at first. And they have a point as, according to a recent study by Nielsen/NetRatings, Amazon.com converts 12.8% of its visitors into sales. Huge number of course and in fact no reasonable guide line for even a multinational concern. In their case a conversion rate of 1% would be a very good value to reach and to hold. The second is the most difficult. It’s not enough to increase the number of visitors, most important is to convince them to do a purchase or to stimulate their desire for more information on products or services.

Therefore, a company should at least consider further investment in their web business to attract more buyers if conversion rate is not satisfying. Afterwards, the value of sales has to be compared to the costs of those actions. When there is an excess a company’s ambitions have been successful. There is no doubt that those investments have to be deliberated on the one hand, but on the other hand they are a great chance to enhance the sales figures. If you know the conversion rate of your web site you can easily calculate the number of visitors you have to attract to win a certain percentage of new buyers.

So why not just check your own company’s conversion rate to get an impression of the efficiency of your web site or just to mark out your position among your competitors? It might not be uninteresting at all.

linked.jpgHow linked are you?

Imagine the ways people arrive on your web site. Besides recommendations of others or non-web advertisement the most effective way most costumers reach a web site via links from other web sites. Do you have an image of the number of pages linking to yours? The characteristic number in this context is the “link popularity” representing the whole number of web sites that link to a single site. Developed by the google search engine founders most of the well known search engines are supposed to use this figure to calculate the ranking for the web sites they provide as a search result. Thus link popularity contributes a lot for position of a web site among the top of the findings.

To get an impression, Amazon.com inbound links reach the number of 18 millions, ebay 9 millions and wikipedia.org even obtains the record level of 42 million incoming links. This can be easily found out by doing a link research with a search engine, in this case provided by google.com. Those levels may obviously not be met by company pages but this topic is all the same very important for them. A high level of link popularity is able to enchance the traffic to a company’s web site and thus this will increase benefits from rising sales. A direct comparison with the main competitors might also be useful to receive an impression whether there is need to work on the own link popularity. A recommendable solution for this survey is offered by marketleap.com which provides listed numbers of inbound links of selected companies. The following excerpt serves as an example for such a research.


Excerpt of a research at marketleap.com platformmarketleap.gif


According to link-popularity.at the philosophy is that important sites will attract many links. Sites with poor content won’t attract many links. This suggests that not all inbound links are handled equally as there are big differences in quality. One possible way to improve link popularity is thus enhancing the quality of one’s web page. Furthermore specialized internet companies offer links for purchase. This method obviously contradicts with the mentioned philosophy but might be effective at all as links of websites with a high link popularity can improve the rank of an other web site. All in all the progression of a company’s link popularity will in every case be reflected in benefits.

top10.jpg How to be on the top 10 of google search results

Who hasn’t already made a google research and wondered why there is always the wikipedia.org site appearing on one of the first ranks. Well, that’s because of all the incoming links to this site. The google PageRank is “a sophisticated method to rank web documents”, according to efactory.de. It is supposed that link popularity contributes in a huge amount to the PageRank algorithm, besides there are of course some other secret factors that also have influence. It was developed by Larry Page and Sergey Brin to bring order to the Web (“The Anatomy of a Large Scale Hypertextual Web Search Engine”).

On the google site itself it says that the PageRank uses the vast link structure as an indicator of an individual page’s value. Every element is being weighted according to its linking structure. The more links refer to a site the higher is the weight of it and weight of the referring page has a huge influence on this effect. The rank is measured on a scale from 0 to 10 with 10 being the highest level, which is reached very infrequently. One can check the PageRank of the own web site over a function plug in in the google toolbar. As an example both Amazon.com and wikipedia.org reach the high level of 8, while BA Mosbach web site only gets a 5 categorization. So what’s your page rank?

As the PageRank has a high economic importance there are a lot of people trying to manipulate the rank by linking to pages with a high level rank. There are even companies specialized in improving sites’ rank by selling high quality backlinks to bypass the algorithm. Thus financially strong companies have easier the possibility to improve. By investing in the quality of the own web site, companies can improve their rank and followingly realize increasement of popularity and success.

lessons.jpg Lessons learned?

Web metrics are obviously not only figures a internet marketer should work with. Every company should be aware of them to be able to appoint the position of their own web site and, if necessary, make improving tactical and strategic decisions. It’s necessary to work with the information one can extract from web metrics. In times where E-commerce is continuously winning importance firms should not only concentrate on their commonplace business but invest in future distribution and information channels in the internet.

And in the end, hopefully you will also be paralyzed by the new opportunities for your company…

reviewer: Anna Martin, Andreas Lothspeich

references:

Advertisements

Comments are closed.

%d bloggers like this: