In a column on ClickZ entitled “5 Traits of the Analytically Empowered Organization,” Neil Mason offers basic guidelines on how to get the most out of measurement and analysis. Mason addresses website analytics but his guidelines can also apply to interpreting press release measurement data collected in Business Wire’s complimentary Newstrak reports.
“In an ideal world, data is integrated around known users but this may not always be appropriate or possible,” wrote Mason. “Some internal data may be on a customer level, but digital data is often based on cookie level data.”
In other words, if you know with whom you are dealing (your own customers, or website visitors about whom you have specific information that they provided upon registration) you can record accurate information about those people. If not, you have to gather information using less direct, and therefore less accurate, methods.
So, although ideally you would like in-depth, accurate data about each visitor to your site, or viewer of your press releases, realistically you might not be able to gather this data without compelling that person to register. But compulsory registration can cause someone to leave a website. A recent study found that 75% of consumers take issue with being asked to register on a website and will change their behavior as a result.
The Business Wire site stopped requiring registration as a prerequisite to reading full press releases because we wanted visitors to stay on our site longer and read more of our clients’ releases. The trade-off is missing visitor-supplied information about their geographic location, industry, job title, and other facts from registration forms.
We can still report upon each visitor’s “location” but that information actually corresponds to the IP address of the visitor. Sometimes that’s the same as where the visitor is located. In other cases, it only reflects where the internet service provider is located.
We’re not alone in swapping broad access and happy website visitors for information that we have to then find indirectly (or selectively, from the relative few who don’t mind registering).
For instance, I’ve seen demos of social media analysis products that have geographic and demographic sections. When I have pressed the salesperson for how these data were derived, I learned they used IP addresses–exactly as we do–and reported demographics for users who include that information in their profiles. In other words, indirect, or selective; or even very, very selective, information.
None of this means that their reports, or our reports, are useless. On the contrary, they can be extremely useful but, as is the case with any statistical report, you must know what you’re viewing.
Here are some tips on how to use statistical reports with these constraints in mind:
- Don’t accept any numbers at face value. Understand the context in which they exist and how they fit in with one another as well as with any statistics you might be gathering internally.
- Make sure you are indeed gathering those internal statistics. Don’t rely on third party reports to tell you the whole story about what you are trying to measure–the “known users” referred to above are YOUR users, and you can leverage those relationships to gather a lot of information.
- Especially if you are purchasing a measurement product, don’t be so lulled by the sexiness of the presentation that you fail to ask the vendor, “How do you find this information?” Prod for specifics so you’ll know what numbers to rely upon and which should be taken with that proverbial grain of salt.