D. Analyzing Data
To understand the customer behavior from the data, the (web) analyst should follow a few initial steps. Following we identify analyses that should help on the conversion of data into insights, which will be essential for optimizing any website.
i. Start from the Basics
Any web analytics tool presents a summary report, a group of basic metrics that are available immediately after logging into the tool. Google Analytics shows the following chart:
Fig. 7: Google Analytics Basic Metrics
The preceding chart shows the following metrics:
Visits: the number of sessions on your website and number of times someone interacted with your ‘site.
Bounce Rate: the percentage of single page view visits (this metric can also have different definitions, such as a visit that last less than 5 seconds).
Page Views number: the number of pages that were requested in all visits.
Pages/Visit: how many pages were seen, on average, in each visit.
Average Time on Site: how long people stayed on the ‘site.
% New Visits: how many sessions were from people who visited your site for the first time.
The preceding numbers will vary from industry to industry, and for this reason there is no absolute benchmark to which a website owner can compare. The best way to proceed is to trend it over time, as much data as possible, to understand if the website is improving or not.
ii. Understand Traffic Sources
Another standard report on Web Analytics tools is the traffic sources report. It usually shows the percentage and absolute number of visitors that came from each type of source. Following is an overview chart as represented on Google Analytics.
Fig. 8: Google Analytics Traffic Sources Summary
iii. Act on the Data, Save Money
In the past, website managers could choose their landing pages for each campaign and have the luxury of deciding how the visitors would start their visits to the website. Today, this control is lost. Search engines decide the website’s landing page. People search (or click through from a link on another ‘site) and go directly deep into the ‘site.
iv. Data Visualization, ‘Site Overlay
Numbers, metrics and spreadsheets are still overwhelming for many; they want to see the data visually represented. The ‘site overlay’ report, or ‘click density’ present in most Web Analytics tools, shows the number of clicks on each link on the page.
Web analysts should look for clusters of heavy clicks, the top two or three most clicked links; then try to reconcile this information against links that s/he wants visitors to click on. S/he should also look at links that ultimately drive high conversions and ask questions such as: do more people convert on the ‘site if they click on product comparison on the home page or go directly to a product page?
It is critical to try to follow the couple of heavy clicks and see what people do next. Walk in their shoes; experience the website through a customer’s eyes.
v. Focus on Outcomes
Most Web Analytics efforts fail to catch on. Few companies are truly data driven. Most people focus on the thousands reports that come out of the Web Analytics tool. Web analysts tend to focus on visits and visitors and parameters and nuances, except outcomes.
As we mentioned in the KPIs section, Web analysts should push themselves to find the “critical few” important metrics for the ‘site. And they are usually linked to the overall objective of the website’s existence. For a blog, it can be the number who visit the speaking engagements page and attend one of the engagements. For a non-profit, it can be to use the core search functionality (e.g. to look for a volunteering opportunity). For an ecommerce website, it is the bottom-line numbers: revenue, conversions, average order value, products sold, etc.
The Web analyst should ask the following questions to be sure to focus on the right metrics:
· Visitors are coming to the website, but is it having any impact on the ‘site?
· If there is an impact on the bottom line, is the website converting enough?
· What’s selling and what is not? Why is it selling? How much of it?
It is fundamental to website survival to understand the customers, this is the only way to understand what action to take on the website to improve and keep pace with the competition.
Advanced Web analytics
Advanced web analytics aims to measure and understand the relationship between the customer and the Web site. Aberdeen Group defines advanced Web analytics as:
Monitoring and reporting of web site usage so that enterprises can better understand the complex interactions between web site visitor actions and web site offers, as well as leverage insight to optimize the site for increased customer loyalty and sales Aberdeen Group (2000).
The field draws from the more general analytic field, which applies complex analysis to large data sets in order to determine value from the information that cannot be achieved through simple means (WordIQ, 2004). They exploit complex statistical data analysis and data management techniques such as OLAP (Thomsen, 2002) to determine complex strategic information. For example, investment and portfolio analytics (Elton and Gruber, 2002) apply analysis techniques to determining the best-matched portfolios for investments, and storage analytics (King, 2002) apply similar techniques to corporate data storage infrastructures to best exploit the resources of a company’s networks.
Advanced Web analytics takes a similar approach
– it is not just about collecting Web site information. The optimum Web analytics strategy couples this information with other data, such as demographics and subscription information, to enable a company to unlock its biggest potential asset – its customers. Optimizing a Web site not only means attracting more customers, it is also about managing existing customers, ensuring the Web site meets their needs and expectations, and turning them from “customers” to “loyal customers” (Whitecross, 2002). In addition, it is necessary to ensure that a Web site also operates in a commercially effective way.
Maximizing the potential of a Web site (i.e. achieving a “successful” Web site), is realized through many different activities. The effective employment of an advanced Web analytics strategy can assist with the tasks listed in Table I, which are all contributing factors to a Web site’s success.
In comparison to the basic metrics, the concept of advanced Web analytics is more centered on a methodology rather than a selection of ambiguous measures.
Metrics and advanced Web analytics
It has been established that what we want to measure is the relationship and interaction between a Web site and its customers. Measurement was previously performed using basic metrics. However, these metrics were contributing to inaccurate and misleading conclusions concerning the success of a Web site. Advanced metrics have evolved as a result of the development of Web analytics as an answer to the shortcomings of basic metrics. Nevertheless, it is not uncommon to find basic metrics incorporated into some of the advanced metrics formulae, as they have proved to be valuable when considered as part of more specific and defined formulae, rather than being used in isolation. These formulae tend to be derived from information requirements on a specific aspect of Web site usage, such as customer lifecycle (Imhoff et al., 2000) or customer behavior. The following outlines these possible areas and provides examples of analytic formulae for each.
The monthly dashboard is a report compiled to assess the performance of the Web site on a monthly basis, over a 12-month period, including visits, visitors, registrations and visits to bookings information. It allows users to compare the activity on the Web site over a year, which permits analysis of the busiest and slowest times over a year.
One must be very careful in making changes to a website, much less to an archival program based on one source of data. Nevertheless, a targeted analysis of the results returned by the Google Analytics report tools yielded several interesting findings regarding how users interact with our site. Initially, we analyzed data from website use in July 2007 and discovered information that spoke to each of our four research hypotheses. After interrogating this data, we identified likely user impediments, redesigned pages to remove the impediments, and made informed decisions to allocate resources toward augmenting particular sections of the website.
Question 1: Which parts of our Website are most heavily used?
This question was answered by the drilling and filtering information found in the Content Drilldown section of the Google Analytics dashboard. By navigating the hierarchy, I was able to determine that 63.2% of the “page views” on our site during July 2007 were in our Archon holdings database, which provides our finding aid system and access to selected digital content.
Question 2: How do people reach our site?
The Google Analytics Traffic Sources Overview provided information regarding this question. As shown in Figure 3, 76% of the approximately 14,000 visits to our site in July 2007 originated from a search engine result page (nearly always Google).32 By contrast, 15.7% of the visits originated when a user entered the URL directly or was placed there when a browser opened the site as the homepage), and 8.6% began when a user clicked a link on another site, such as Wikipedia or our parent institution’s website.