What works and what doesn’t in online advertising + key benchmarks

MarketingSherpa’s 2008 Online Advertising and Benchmark Guide + Benchmarks was born out of a belief that online adverting at present is misunderstood, and as result, underutilized. This isn’t to say that existing advertisers should simply increase spending online. Rather, we think advertisers that can find a balance between economical, efficient targeting and clutter-busting, highly engaging advertising will achieve far better ROI for themselves and a much more positive online experience for consumers. Many moving pieces fit into an online campaign, so this is no easy task.
The first hurdling block, which must be overcome, is that old-school media math is based on the limitations of traditional, linear, analog media, and doesn’t reflect the realities of non-linear, digital media. For this reason, digital media math needs to get more sophisticated, and more accepted among media professionals. One of the more obvious ways of adding sophistication to digital media buys is to take a closer look at frequency. Traditional media doesn’t allow for frequency to be controlled on the individual level, but digital does. To that end, we obtained data from InsightExpress showing how frequency affects ad effectiveness, we looked at aggregate conversion rates by frequency of exposure from Doubleclick, and then explained how to implement a frequency capping policy that makes sense for each advertising strategy.
Another limitation of traditional media planning and math is a lack of consideration for quality on the level of the individual consumer. With traditional buying techniques, a certain amount of “waste” is inherent and difficult to account for. Digital advertisers can and should be factoring in qualitative metrics when planning media. This can take many forms, from advanced behavioral targeting to assigning value by conversion rates. We showcase ways of assigning quality to placements through eyetracking, and effectiveness of media through cross-media effectiveness studies. The point is that calculating effective reach rather than just reach should be the norm on digital platforms.
We don’t think there is a magic bullet for creating the perfect ad, and really encourage people to get creative and try new things. According to our research, it’s the advertisers that try new things and constantly test them that consistently do well. We strongly encourage research and testing and show proof from our survey that qualitative research, which affects the insights going into ad creation, can actually be more effective from an ROI standpoint than improving tracking or A/B testing.
Finally, Analytics needs to get better at incorporating modeled metrics for effectiveness. By designing branding dashboards that incorporate both brand metrics projected from survey sample data with observed, tracked metrics like impressions and clicks, it’s possible for marketers to get a fuller picture of what’s actually happening with an online campaign. There’s too much data and not enough insight out there.
We don’t have all the answers, but we do have a lot of them, and where we don’t, we hope to provide discussion, new ideas, and testing. Pushing online advertising from where it is to where it could be will be a slow process, but it’s one we look forward to participating in.