Sometimes it makes sense to track individual visitors to your site and create a feedback loop.
There are three kinds of unique visitors coming to your web site—truly anonymous, mostly anonymous, and known visitors [Hack #5] —and each category affords you more or less detail about the individual behind the clicks. The latter category, “known” visitors, has a great deal of potential for some business models to use web measurement to impact both the online and the offline relationship with the visitor. Smart web data analysts will go the extra mile to report about these visitors and create action plans based on their usage patterns.
Briefly, the known visitor is someone you can positively identify in the offline world, either by virtue of a unique login or some other authentication scheme. The common log format (CLF) has a field for authenticated usernames:
127.0.0.1 - frank [10/Oct/2000:13:55:36 -0700] "GET /index.html HTTP/1.0" 200 2326 - "Mozilla/4.08 [en] (Win98; I ;Nav)"
The frank
in the preceding
server logfile entry indicates that someone has successfully logged
into a domain requiring authentication as “frank"and then made a
request for the file index.html. Solutions based
on JavaScript page tags also commonly accept a unique user or customer
identifier:
hbx.hcv="";//CONVERSION VALUE hbx.cp="null";//LEGACY CAMPAIGN hbx.cpd="";//CAMPAIGN DOMAIN //CUSTOM VARIABLES hbx.ci="epeterson93111";//CUSTOMER ID hbx.hc1="";//CUSTOM 1 hbx.hc2="";//CUSTOM 2
In this example, the JavaScript declaration hbx.ci="epeterson93111"
; identifies a
visitor in real time as epeterson93111
to WebSideStory’s HBX code.
The critical aspect in identifying a known visitor, one outside of the
scope of this hack, is connecting the unique user identifier (UUID) to
a database of personal information about the users. The most common
personal information connected to web measurement data are things like
name, company, title, department, phone number, and email address. As
you can see in Figure
4-12, this type of information can then be used to build useful
reports about the individual (but don’t forget to read [Hack
#26] before you do anything
like this!)
Remember, the known visitor is someone who is identifying herself to you, either because she’s required to do so to use your system or because you’ve provided her a significant incentive to do so. The former group is often composed of intranet/internal users or business partners logging into a secure extranet, the latter group is often made up of customers of some kind. Table 4-1 provides a handful of examples of internal/business-to-business (B2B) and external/business-to-consumer (B2C) visitors that are easily identified.
Table 4-1. Examples of know visitors that you may want to measure
Internal/B2B visitors | External/B2C visitors |
---|---|
• Sales and support staff | •Registered customers |
•Executives | |
•Business partners | •Identified leads and prospects |
•PR and marketing agencies | •Auction participants (e.g., eBay buyers and sellers) |
•Outside graphic design agencies | •Newsletter subscribers |
•Business-to-business customers | •Registered visitors |
While this list is far from exhaustive, hopefully you get the idea. Once you identify the known visitors you want to track, the next step is to figure out which reports you’ll generate and what you’ll do with those reports!
In general, there are four categories of reports useful to helping you better understand your known visitors: activity reports, content consumption reports, customer support reports, and sales analysis reports. Each of these types of reports is designed to provide you with a wealth of information to understand customer needs or proactively market.
The most common report type for intranet users, activity and application usage reports, can help you understand who within your organization has adopted the tools you’ve built for your intranet and who is likely using the wrong resources (for example, bugging other people when they could be self-servicing via the intranet). A great strategy for increasing intranet adoption in organizations is to sort departmental usage by employees and have the top 10 percent of users offer intranet training to the bottom 25 percent of users!
Content consumption reports are similar to activity reports, but generally focused more on external known users in an attempt to identify preferences for products or content. Once identified, these visitors can be segmented and directly marketed to in an effort to increase loyalty and future conversion. A useful strategy is to generate a list of visitors who spend more than 10 minutes browsing information on a weekly basis and communicate to them differently than you would a less engaged visitor.
Usually, a report describing how visitors interact with your customer support content will help make your customer support agents smarter and more helpful. If you’re able to generate this information, why wouldn’t you show your support agents the list of articles and pages the customer was looking at prior to picking up the phone so that he can better assess the likely problem? A must-generate report for any company dedicated to providing highest-quality support to their customers!
When retailers get really good at outbound marketing and up-selling to existing customers, a sales analysis report can help you mine for likely prospects. The best example is probably customers who have purchased a particular product and who have looked at but not purchased ancillary products. Sending those folks a “We know you already have product X and we’re having a sale on widgets for product X this week only” often generates incremental sales.
Some web measurement application vendors, especially those providing hosted solutions, don’t include anything but the most rudimentary “who visited” reports for known visitors. Still, you may want to understand activity for these visitors and get you own metrics like page views, visits, and unique visitor counts. A clever hack for doing this is repurposing your content group variable to include the known visitor’s UUID. For example, use the following code to identify my page names and content groups via a JavaScript page tag:
… ' ALREADY IN A JAVASCRIPT BLOCK FOR THE PAGE TAG var _pn=document.title; ' SET PAGE NAME (_pn) TO THE HTML <TITLE> var _mlc="/" + "CATEGORY" + "/" + "SUBCATEGORY"; … ' JAVASCRIPT CONTINUES
To this code you can simply add a userID
variable before the content group to
capture high-level information about individual users:
… ' ALREADY IN A JAVASCRIPT BLOCK FOR THE PAGE TAG var _pn=document.title; ' SET PAGE NAME (_pn) TO THE HTML <TITLE> var _UUID="USER_IDENTIFIER_IF_KNOWN
"; if (_UUID != '') { _userID = "/" + _UUID; } else { _userID = "/anonymous_visitor"; } var _mlc=_userID + "/" + "CATEGORY
" + "/" + "SUBCATEGORY
"; … ' JAVASCRIPT CONTINUES
This way, you’ll end up with a report that has, at the top
level, a list of usernames (or user identifiers) and a group for
anonymous_visitor
with the
appropriate metrics associated with each. Ideally you’ll then be able
to drill down into any of the users and see what pages and content
they viewed.
Warning
While this is a pretty good hack for tracking known visitors, never do this in your main traffic measurement account. Doing so will hamper the normal reporting process significantly. Instead, ask your vendor for an extra account for this kind of tracking and get their help setting it up if necessary.
While this may seem like a lot of extra work, it’s important to consider the easiest person to sell to or connect with is somebody you already know. Every successful business person knows this, so don’t be afraid to measure the known visitor and take advantage of the relationships you already have. Be respectful and don’t be intrusive, but take advantage of this powerful strategy.
—Akin Arikan and Eric T. Peterson
Get Web Site Measurement Hacks now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.