Wednesday, August 6, 2008

E-marketing Effectiveness Benchmarks for the Pharmaceutical Industry

ComScore has released its 2nd Annual e-Marketing Effectiveness Benchmarks for the Pharmaceutical Industry. The study was based on passive observation of ComScore's million person panel across 30 drug brands. Some very interesting metrics that have hitherto been unpublished or unavailable on the effect of e-marketing on script uplift and incremental next fill/adherence are revealed in the study.

According to the study, 90 million Americans visited health-related Web sites in March 2008. This is a 32 percent increase in two years, a rate of growth over 3x the general Internet growth rate of 10% over that same period of time.

U.S. Unique Visitors - Total Internet and Health-Related Sites

US Internet Users Who Have Searched Online for Medical Information
Source: eMarketer

Top 10 US Health and Medical Web Sites March 2008Source: eMarketer

The ComScore study measured incremental effect on aided and unaided awareness and brand favorability, looking at online banner ads, interactive rich-media ads, and visits to a branded web site or unbranded "" site. Perhaps not surprisingly, online advertising effectiveness increased with richer interactions, rising to its peak with visits to a Web site. Unaided awareness from visiting a Web site and brand favorability rose significantly to a high of 36% and 28% respectively with prospects, those not using the drug in the year prior to the study, and favorability rose 15.8% among patients.

Throughout the report no mention is made of abandonment rate's (visitors who leave a site after viewing just one page) effect on the results, although the study does discuss the importance of number of pages viewed in general. In any case, it is a truism for measuring Web site effectiveness (and is a standard measure used by all Web metrics programs) that increasing number of page views increases visitor engagement with a site and thereby also any of the benefits from that visit. The study also does not state whether length of time on site, another standard Web site effectiveness metric, showed any significant uplifts. Presumably not, or it would have been mentioned.

Things start to get very interesting in the next two charts where data is presented that could lead to some hard ROI calculations for pharmaceutical Web sites. These are the first published metrics of script uplift and incremental adherence/next fill from online advertising that I am aware of. Among new patients the study showed an almost 5% script uplift from prospects visiting the "" Web site and an almost 20% increase in adherence/next fill among patients visiting the Web site.

Again, the report does not discuss whether site abandonment was factored into its results, but one could always measure both single page visits and multipage visits to calculate an ROI range from conservative to optimistic. It is very difficult to know which visitors to a "" Web site are prospects and which patients, so while these metrics are an excellent start, they still fall short of a methodology to calculate a hard ROI for pharmaceutical Web sites. Presumably ComScore asked the members of its panel who participated in this study whether they were users of the drugs being studied to determine who was a prospect and who a patient. This is a luxury that is more difficult to realize in the wild, although incentives can obviously be provided to visitors to part with anonimized personal information. From this one could periodically calibrate the percentage of prospects to patients at a "" Web site and from this have the beginings of calculating a hard ROI based on increase sales of the drug.

It is interesting that online ads showed no effect on script uplift among prospects, but showed a 4% - 9.5% increase in next fill/adherence among patients. Perhaps this is like a mother's reminder to take your medicine: Ignored if one is not already doing so, but otherwise a word to the wise is sufficient. This also suggests a more targeted approach for these ads towards patients. (Something like - "Don't forget to take your medicine" -- Mom.)

For the first time ComScore was able to measure incremental conversion by referral source: Display media, search (both paid and algorithmic), and organic (or direct navigation). For prospects, conversion to a new sale showed an almost 12% increase over the control group with direct navigation to a "" site. This was 3x - 4x higher than the incremental conversion from search and display media. This is a very significant finding. I would have expected the large difference over display media but would have expected a much smaller difference with search, especially algorithmic search. This suggests that cross media campaigns, viral and word of mouth campaigns, and social networking-focused campaigns that build brand awareness and generate direct navigation to a site, rather than SEM or SEO, are the best way to create script uplift online.

Incremental adherence/next hill uplift for patients was virtually the same across all of the media types. Again, a word to the wise is sufficient no matter the source.

The study is available upon request from ComScore.

No comments: