There is hardly a topic that is reported on and discussed as nebulously as the reach of online media in the wine sector. Different metrics from different measuring tools are wildly jumbled up here. It gets even more adventurous when online media are compared with print media. But more on that in a later blog post. Today I'll stick to online media.
In the online industry there are clearly accepted standards for measuring the distribution of online media. All the big players, agencies and Webpages use clear and unambiguous terminology. This can be discussed, is discussed and will continue to develop. But the status quo is that these standards exist and there is no reasonable reason not to adhere to them.
For years I have been trying to convince online colleagues in the German-speaking wine sector to use standards and ideally a transparent, commonly used measurement procedure. If we want to be taken seriously as an online medium, we also have to communicate comprehensibly and on the basis of serious figures. Unfortunately, my efforts have so far come to nothing. Fear of transparency?
Page impressions: The number of pages that people view on a website in a given period of time. More information in the glossary
Visits: The number of visits by people using a website in a given period of time. If the same person uses a website more than once, each use counts as a new visit. More information in the glossary.
Unique Visits: The number of unique visits to a website. Multiple visits by the same person only count once..
Length of stay: The total amount of time people spend on the website, regardless of how many pages they view.
All other metrics or measurement methods are fuzzy, self-invented, uncommon and do not serve a serious estimate of a website's reach, nor do these numbers serve a comparability within the industry or with other industries. Nevertheless, Technorati, Wikio, logfile analyses, Google-Pagerank and other sources of figures enjoy great popularity. Freely according to the motto: I communicate where I myself perform best.
There are many methods to measure the above figures. One must distinguish between two basic methods:
Log file analysis: Inaccurate, unprofessional and strongly falsifying in the result. Often higher by a factor of 10-15 than would be correct
Pixel analysis: The usual, professional method for measuring the reach of a website. Common tools that use pixel analysis are e.g. Google-Analytics, Etracker, the IVW partner and many other professional service providers. The results of all service providers differ slightly, but the order of magnitude of the results is the same.
A more detailed description of the differences can be found in the glossary.
How credible and controllable are the figures communicated by website operators? Hardly anyone gives users transparent access to the respective measurement system. The IVW (Informationsgemeinschaft zur Feststellung der Verbreitung von Werbeträgern e.V.), which also measures and publishes very credible and reliable reach figures in the print sector, would be ideal here. With its partner InfOnline, the IVW offers a serious measurement based on pixel analysis. Unfortunately, this is relatively expensive for website operators, and support is so always on hand that it is unlikely to catch on.
But there are good ways to estimate the numbers yourself:
Both methods are based on sampling and are imprecise; however, they are a rough guide.
Originally, I wanted to list all known wine websites and the results for the reach in Google and Alexa at this point. For this purpose, a comparison with the figures communicated by the providers themselves was planned. However, the discrepancies are sometimes so hair-raising that I would like to dispense with this comparison and, in two cases, also with naming names. I simply credit those concerned with having lost their way in the confusion of terms and numbers themselves.
Here is a summary of the results for the different segments:
All figures are rounded and were retrieved from Google and Alexa on 2.2.2011 and refer to one month, unless otherwise stated.