The corporate Web site. Every company has one; and at some point, the marketing department has attempted to use online advertising as a means of attracting prospects and ultimately helping fill their company's sales pipeline. This is in a business-to-business (B2B) context, of course, where the Web site typically plays the role of online brochure, demo center and self- service customer support representative.

Yet for the tens of thousands of B2B Web sites, online promotion is not as popular as it once was. In July of 2002, research firm GartnerG2 reported that in 2001, B2B marketing efforts comprised less than 11 percent of all online advertising dollars spent. The reason? The promise of reaching highly targeted and motivated audiences via online advertising often went unfulfilled. In an era of declining marketing budgets due to an overall decline in the economy, most businesses simply cannot afford investing in such unproductive activities.

How do businesses know whether or not online marketing is helping them reach and attract the right audiences? Web analytics, that's how. By now, anyone who has worked in marketing over the past several years has seen a report of Web site activity. Whether these reports have come from WebTrends (now a division of NetIQ) or any of the numerous application service provider (ASP) offerings now available, the insight into promotional effectiveness presented by such products can be very powerful. Yet, with fewer B2B Web sites investing in online promotion, how are marketers supposed to gauge the value of their Web site investments? Certainly not by using the simple hit/click/visit statistics of the past.

What's needed instead is a new form of Web analytics that helps businesses understand the extent to which the Web site supports key offline business processes, particularly those related to customer relationship management (CRM) efforts. In this manner, a direct connection could be established between the Web site and business performance – ­a concept otherwise known as Web site visitor intelligence.

Web Versus Business Process Viewpoint

The information captured by typical Web analytics products and services is usually presented in a Web-centric manner –­ visitor data is made available in logical groupings such as page views, referring Web sites and entry pages. Some companies that serve this market have come up with ways of combining data into metrics that lend greater value to an analysis (such as the conversion rate by which a visitor enters a site and then goes to a desired destination within the site). More advanced offerings enable businesses to tie customer information to Web site visitor data so that reports and analysis are based not on anonymous visitors, but on people's actual names and demographics.

What these offerings fail to capture, however, is the extent to which this activity supports the company's overall business performance. For example, don't marketers and other stakeholders in a business have a vested interest in knowing what prospects, customers, competitors and job recruits are visiting the Web site and what content they are viewing for how long and how often? There are a number of functions represented on most B2B Web sites, yet the professionals responsible for those functions often have no clue about the extent to which the Web site helps them, if at all. Some common questions you might expect include:

  • From a Sales Manager –­ For prospects in my pipeline, what accounts viewed the Web site this week, and what did they look at? I'd like to know their top concerns so I can position us more effectively in the account. Can I also see company demographic information, such as revenue and number of employees, or if they've been a customer in the past?
  • From a Marketing Manager –­ What online marketing programs drove leads into the sales pipeline? What competitors visited the site this week, and what content did they look at? What university recruiting programs have driven students to our area?
  • From a Customer Support Manager –­ Which of our most valued customers sought information from our support Web site today, and what are they seeking? I'd like to be proactive and contact them if it appears they are looking for, but not finding, answers.

These are all relatively simple CRM-oriented business questions that could be answered based on the data collected by traditional Web analytics products and services. However, there is no simplified or automated means by which managers can get the answers. The vast majority of businesses have neither an expensive operational CRM system nor a staff of analysts to compile and routinely distribute reports.

Flipping the Web Analytics Paradigm

To draw the critical connection between Web site activity and offline CRM business processes, what's needed is an extension to current reporting solutions that relates visitor behavior to desired business outcomes. Viewed as a scorecarding framework for business processes such as a sales cycle, this analysis could show, for example, that visitors from a prospect in the sales pipeline sought product information during the previous week. In isolation, this has little analytical value; however, if you are the salesperson and the company in question is in your pipeline at a near-closing stage, this information can be very useful. It would be equally important to know if a key competitor in the deal had been scouring the Web site over the same period of time. The issue then is how to present this insight in a timely fashion and in a manner that makes sense.

At the most basic level, what I am talking about is flipping the Web site activity reporting paradigm on its head. In other words, rather than viewing Web activity according to how a visitor browses the Web site, understand the contribution of site activity to a business process such as a sales cycle, marketing program or customer support effort. For example, sales managers should be able to see prospects in their pipeline that visited the Web site, as opposed to a list of referring visitor domains, some of which may be prospects. Likewise, customer support managers should be able to view the most valued customers who spent an hour per day on the support site versus visits by authenticated users sorted by domain, only some of whom may be customers. See the distinction? One of the key reasons Web efforts are often perceived as a necessary expense (but with questionable return) is that few businesses can draw a connection between Web activity and the day-to-day offline business processes that preserve and drive new revenue.

The simplest manner in which to accomplish this is to visually map visits to the business process in question according to referring visitor domains ( Given that most B2B Web sites attract visitors from corporations, this is a reasonable assumption, relative to the thousands of anonymous and indistinguishable visitors a business-to-consumer Web site might receive from users of, for example.

Were this a commercially available product, it would allow a business user to define natural language "entities" to monitor within a business process context ("entity" being a company or person, identified either by authenticated name or company domain name). Better still would be connectors to the source systems (CRM, sales force automation, accounting) to automate entity configuration. Entities might include competitors, customers, prospects in the pipeline or universities subject to recruiting efforts. These definitions would then seek matches to probable referring visitor domains (or authenticated users) within the Web site activity database, and a scorecard would then be published for review by the user, ideally sent to their personalized enterprise portal or e-mail box.

Beyond the Basics

While visitor intelligence scorecards account for known (customers/competitors) or expected (prospects) entities, what about the many more unexpected visitors that routinely hit the Web site? Additionally, what about the contextual information that can add tremendous value to an analysis, such as a customer's history or prospect's demographic profile? What about being able to segment visitors by industry to determine if marketing is attracting the intended audience?

To address the first issue – ­ accounting for unexpected visitors ­– data and text mining concepts could be applied to create profiles of the types of firms visiting the site. This would enable the marketing department to understand whether or not online and offline marketing programs are attracting the intended audiences, as well as help distinguish between visits from potential sales prospects, competitors, job recruits or casual visitors.

In practice, a text mining-enabled site crawler would be deployed against referring visitor domains on a real- time or batch basis to scan meta tags, key words and site text to make determinations as to whether visits stem from probable customers, active prospects, suspects, competitors or job recruits. Data mining would then segment visits by industry, competencies, partnerships, sales offices, customers and other details that provide contextual information to visitor intelligence scorecards. Importantly, this would also enable the company to develop a genuine information asset that might be employed in an operational CRM/Web site personalization platform.

In addition to looking outward to visitors' referring domains, visitor intelligence scorecards could obtain data from subscription information services, such as Hoovers. For salespeople, this would allow them to see important background information associated with prospects such as revenue or number of employees, or enable marketing people to see details of organizations identified as probable competitors.

Thus far, visitor intelligence scorecards have focused on using Web site activity data to illustrate the site's contribution to various CRM business processes. However, without access to multichannel internal information, scorecards lack key contextual information that can enhance decision support.

While the referring visitor domain crawler looks beyond the Web site to understand unexpected visitors, a similar concept ("enterprise information crawler") could be deployed internally to make scorecards aware of relevant corporate information. Using text mining technology, enterprise data sources could be crawled to create a logical mapping of data sources such as customer histories, sales pipeline accounts and financial system information. With this crawling complete, scorecards could provide contextual information "on demand." Sample applications would include a customer service manager instantly seeing support records associated with high-value customers spending significant time on the support Web site or salespeople knowing if a new prospect has been a prior customer.

B2B marketers have traditionally relied upon online advertising and promotion as a means of attracting prospects and ultimately proving the value of Web efforts to their organization. Yet online promotion is less popular among B2B marketers than it was in the past. To justify Web site investments, marketers and companies in general need new solutions, such as visitor intelligence scorecards, that help managers in all functional disciplines understand the extent to which the corporate Web site supports their offline business processes.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access