Sibel Satiroglu, presesenting at eMetrics Washington (http://emetrics.org/washingtondc/) Presentation started. "Answer the question: What happens after the click?"
To identify the QUALITY visitor, need to understand the Post Click behaviour beyond CPC and CTR.
Focus on improving the micro-conversion to yield higher macro conversion (sales). Micro-conversion is important because the majority of sales are not achieved on HP's website.
Must Have Metric #1: Abandonment rate is a critical metric. Abandonment = absolutely no clicks on the site. Abadonment metrics is a 'consideration metric'.
Must Have Metric #2: Conversion metrics are 'preference metrics'. Can be download a case study as well as buy now.
Nice to have metric #1: 1st time HP visitor, #2: Visit duration. Helps speak the language of the executives. Don't use average time spent because it is dominated by extremes. Use 75th percentile visit duration to distinguish a 'very engaged' visitor.
Numbers aren't meaningful. Meaning comes in comparison. [ Absolutely! ]
Cost metrics: CPC plus cost per engaged visit' and 'cost per conversion'
Case study #1: Print payback campaign - currently running. CTR and CPC metrics looked very promising but the post click metrics told a different story. Placement targeting - almost all abandoned. Contextual targeting 8/20 ad groups had super high abondonment rate. Action taken: Stopped all placement targeting, scaled back contextual.
Case Study #2: Contextual targeting complements standard search, drawing new highly engaged visitors (75th percentile duration). Compared to std search, contextual generally yields similar abandonment rate and cnvn rate, more new visitors, higher duration (up to 2x)
Case Study #3: Consumer creative studio - 25-30% traffic driven by paid search at a low cost per print. Prints are monetized as future ink/paper sales.
This was a lesson in relevancy. Top keywords were valentine cards. Solid CTR, Low CPC. Post click. Print cnvn 4x below, cost per print 6x above seasonal norms. Site did not have cards. Had other Valentine images, prints, but not cards.
Actin: redircted 1000 kw to deeper landing ages for increased relevancy between search intent
Result: Achieved 6x higher print cnvn and 17x lower cost per print vs Valentine's day.
Case Study #4: Supplies Message Testing. Which message is most appearling to the searcher.
Ink - 65% more had significantly higher CTR. Had 1.7x lower abondonment 10 % more conversion and lower cost per print.
Findings aligned with qualitative research results. Statement scored as most interesting, most believable, most likely to consider, would influence future buying decisions. Sharing this qualitative information helps bring more traditional marketers on board; increases credibility in their eyes.
What about the Macro Conversion? Most converting offline. Therefore, HP hasembarked on research project with Google to determine the effect of paid search on in-store sales. Will go live in January. Will saturate test local market with 100% paid search impression share.
Also an opportunity to improve cross-channel experience.
Questions...
Q: How do you differentiate between consumer and enterprise behaviour? A: Enterprise difficult to reach. Set benchmark of abandonment higher for them. e.g. abadonment hurdle may be 50% for consumer (anything above is bad). Set abandonment at 70% for Enterprise,
Q: About the pilot with Google. What's are you looking for. A: This is a research pilot. Looking to find whether there's a small impact or large impact.
Q: is placement targeting generally worse A: Case study present was placement targeting for enterprise. Have only tried it 2x. Both did not perform post click.
End session. Thanks Sibel!
june_li
Copyright 2012 Scribble Technologies Inc. All Rights Reserved.
