I first wrote an article on CPW in 2001. We had been using this term at Mediasmith for several years and thought that we needed to put a stake in the ground on the term. Since that time, it has been used by select people in the industry but only recently has gained broader acceptance as the industry strives for metrics that can be understood by all.
What is CPW? It is simply "cost per whatever" the advertiser is basing their ROI on. This can be cost per sale, cost per download, cost per sign up for a newsletter, whatever metric means success. It could be an action such as registration or purchase; it could be a passive measure such as pageviews or unique users. Cost-per-pageview has proven to be an especially effective measure for media whose whole model is selling advertising on their site. In effect, they make money on the spread between the difference in pageviews that they sell and pageviews that they buy.
CPW can also apply to more sophisticated metrics. For instance, a client may wish to evaluate a visitor not on their first visit, but may want to try to emphasize sites and creative that brings back visitors multiple times. We had one client that believes that visitors are only valuable if they come six or more times to the site. In this case, we would measure repeat visits and place future dollars on sites that do the best job of bringing consumers who are frequent visitors. If a confirmation page on a web site can be dedicated to the action, this action can be the "W" in CPW.
We all know (or should know) by now that CPC is dead. Clickthrough rates, once the hallmark metric of Web success, have proven to be irrelevant and it is time we moved on. Actually, they have been irrelevant for years but the success of Google has made this metric pervasive. The web has taught us however that the old method of computing what media costs does not make a lot of sense. What does make sense is how much it costs to drive success or ROI.
Time-tested analysis has established that there is NO relationship between clickthrough rates and success of a Web campaign. There are two related factors here: 1) just because someone clicks on a banner or link does not mean that they do anything on the site they are sent to. In fact, the correlation of clickthrough and back-end action of any sort varies all over the lot. 2) An increasingly large percentage of consumer reactions to a Web campaign are from those who are exposed to a banner but do not click and later visit a site. While the Web average is in the 50% range, this number rises to higher than 50% when traditional, offline media are employed at the same time as Web. Or when a campaign is more branding oriented. Or when the content is especially "sticky" and people don't want to immediately click off of the content to go somewhere else. This concept, called view through or post impression is not embraced by all. They point to PR or other media factors which could have driven the consumer to the site. All the same, there is a direct data correlation that can be established between those who see an ad and later visit the associated content. Yes, the more activity outside of the web that is included in the campaign (e.g., traditional media), the higher the view through. A "surround sound marketing" campaign increases the chance for view through rather than post-click. At Mediasmith we've had a number of campaigns where the overwhelming majority of conversions took place without a user clicking on an ad. Also, it should be noted here that the longer someone has been on the Web, the more likely they are to be a post-impression rather than a post-click visitor.
We are not suggesting that CPW be a basis for payment. Although we do expect that some sites will offer this as back-end tracking becomes more sophisticated. There is a new generation of tagging being tested that could not only eliminate the need for third party ad servers but could give the seller as well as the buyer visibility on view through and associated conversions. CPW should, however, be the basis for evaluation of sites and creative. After all, isn't it all about the advertiser and ROI?
Now we recognize that the concepts above are not necessarily new but could be called simply a new label on something old. But sometimes names are everything in achieving clarity in communications. Some people call this method CPA. But since CPA is also a method of selling, we find that the use of CPA to describe back end evaluation sends the wrong message in analytics wherein some readers of a report do not understand that the underlying buy is, in fact a CPM buy.
Lastly, the use of the term CPW injects a little smile when someone first hears it. And any analytics conversation can use that.
