There are a bewildering range of hundreds of web analytics tools, varying from shareware packages with often primitive reporting through to complex systems which may cost hundreds of thousands of dollars a year for a popular site. Given this, it is difficult for the Internet marketer to select the best tool or tools to meet their needs. One of the first issues to consider is the different types of measures that need to be integrated within the performance management system. Figure 9.6 gives an indication of the types of data that need to be integrated; these include:
Data would be ideally collected and reported within a single tool at this level, but unfortunately to obtain the best reporting it is often necessary to resort to four different types of tools/data source:
• Referrer data from acquisition campaigns such as search marketing or online advertising. Separate tools are often also required for retention e-mail marketing.
• Site-centric data about visitor volume and clickstream behaviour on the web site.
• Customer response and profile data.
• Transactional data about leads and sales which are often obtained from separate legacy systems.
2 Tactical data
These data are typically models of required response such as:
• Reach models with online audience share data for different demographic groupings from sources such as Hitwise and Netratings.
• Lifetime value models which are created to assess profitability of visitors to the site from different sources and so need to integrate with operational data.
Performance management systems for senior managers will give the big picture presented as scorecards or dashboards showing the contribution of digital channels to the organisation in terms of sales, revenue and profitability for different products. These data indicate trends and relative performance within the company and to competitors such that the Internet marketing strategy can be reviewed for effectiveness. The efficiency of the processes may be indicated, through, for example, the cost of acquisition of customers in different markets and their conversion and retention rates.
So, an important requirement of a web analytics tool is that it should seek to integrate all these different data sources. The other main requirements of a web analytics tool to consider include:
• Reporting of marketing performance (many are technical tools which do not clearly report on results from a marketing perspective);
• Accuracy of technique;
• Integration with other marketing information systems (export);
• Ease of use and configuration;
• Cost, which often varies according to site visitor volumes and number of system users;
• Suitability for reporting on e-marketing campaigns.
Many online tracking tools were originally developed to report on the performance of the site and the pages accessed rather than specifically to report on e-marketing campaigns. It is therefore important that companies have an excellent campaign reporting capability. When online marketers are reviewing the capability of tools, they should be able to answer these questions:
1 Can the tool track through from point entry on site through to outcome? For example, to outcomes such as registration, lead or sale? Integration with data to reflect actual leads or sales in a legacy system should also be reported.
2 Can the tool track and compare a range of online media types? These were explained in Chapter 8, for example, interactive (banner) ads, affiliates, e-mail marketing, natural and paid search.
3 Can return-on-investment models be constructed? For example, by entering costs and profitability for each product?
4 Can reports be produced at both a detailed level and a summary level? This enables comparison of performance for different campaigns and different parts of the business.
5 Is there capability to track clickthroughs at an individual respondent level for e-mail campaigns? This is important for follow-up marketing activities such as a phone call, direct mail or e-mail after an e-mail list member has expressed interest in a product through clicking on a promotion link.
6 Are post-view responses tracked for ads? Cookies can be used to assess visitors who arrive on the site at a later point in time, rather than immediately.
7 Are post-click responses tracked for affiliates? Similarly, visitors from affiliates may buy the product not on their first visit, but on a later visit.
8 Do e-mail campaign summaries give unique clicks as well as total clicks? If an e-mail communication such as a newsletter contains multiple links, then total clicks will be higher.
9 Is real-time reporting available? Is immediate access to campaign performance data available (this is usually possible with browser or tag-based campaign tracking solutions)?
10 Are cross-campaign and cross-product or content reporting available? Is it readily possible to compare campaigns and sales levels across different products or different parts of the site rather than an aggregate.
Accuracy is another important aspect of web analytics tool and managers need to be aware of some of the weaknesses of web analytics tools based on log file analysis. Perhaps the worst problems are the problems of undercounting and overcounting. These are reviewed in Table 9.2.
PERFORMANCE MANAGEMENT FOR INTERNET MARKETING Table 9.2 Inaccuracies caused by server-based Log file analysis
Sources of undercounting
Sources of overcounting
Caching in user's web browsers (when a user accesses a previously accessed file, it is loaded from the user's cache on their PC)
Caching on proxy servers (proxy servers are used within organisations or ISPs to reduce Internet traffic by storing copies of frequently used pages)
Firewalls (these do not usually exclude page impressions, but they usually assign a single IP address for the user of the page, rather than referring to an individual's PC)
Dynamically generated pages, generated 'on the fly', are difficult to assess with server-based log files
Frames (a user viewing a framed page with three frames will be recorded as three page impressions on a server-based system)
Spiders and robots (traversing of a site by spiders from different search engines is recorded as page impressions. These spiders can be excluded, but this is time-consuming)
Executable files (these can also be recorded as hits or page impressions unless excluded)
A relatively new approach to the problems of undercounting and overcounting of server-based log file analysis described in Table 9.3 is to use a different browser-based or tag-based measurement system that records access to web pages every time a page is loaded into a user's web browser through running a short script, program or tag inserted into the web page. The key benefit of the browser-based approach is that potentially it is more accurate than server-based approaches. Figure 9.7 indicates how the browser-based approach works.
Server log file
Was this article helpful?
Co-op Mailing means that two or more businesses share in the cost and distribution of a direct mail campaign. It's kind of like having you and another non-competing business split the cost of printing, assembling and mailing an advertising flyer to a shared same market base.