|
While planning a recent dinner party, I considered putting homemade gnocchi on the menu — a first for me. My cookbook said I needed a potato ricer, which I didn't have in my kitchen or even lexicon. So I opened my laptop, searched for “potato ricer” and began my fall through the purchase funnel.
While doing so, I took note of a few insights on search. They include the following:
1. Search is at the center. The purchase funnel really isn't a funnel but more a cycle — and search is at the center. From awareness to consideration to purchase, search links each of the phases. In my case, I first searched Wikipedia to learn about potato ricers, searched Google to find brands and then Google again to find a specific retailer. One implication of this cycle is that the line traditionally separating brand and direct marketing is less clear. As such, marketers should adopt more cohesive online strategies. This means, for example, combining display and search advertising.
2. Always be closing. In search advertising, keywords typically enjoy a disproportionate amount of attention and spend. While certainly important, keywords represent only one side of a sale, the “pitch.” But landing pages represent the other side, the “close,” and are just as important. A retailer’s search ad for “potato ricer,” for example, should land on a page for a specific product or list of products, not a homepage.
Furthermore, the product shown should be the one most relevant to the consumer: A 29-year-old male browsing at 1:30 a.m. isn't likely to spend more than $40 on a potato ricer, so he shouldn’t be shown a $50 one first. Long story short, I didn’t buy a ricer.
3. Data is king. That online channels — search, display, contextual and affiliate — don't exist in silos should be clear. The correlation between channels is less straightforward, however. To solve this attribution and allocation problem, marketers should rely on data and analytics. Data exchanges and ad servers offer a mother lode of cookie- and impression-based information begging to be mined and analyzed. The same holds for landing pages, which can be optimized in a number of ways, including multivariate testing or predictive modeling. Marketers can improve the effectiveness of their campaigns using these approaches.
Brad Berkin is a marketing analyst at [x+1], a New York-based digital media and site optimization provider for agencies and advertisers. He can be reached at bberkin@xplusone.com.
Companies Mentioned:
 |
|
Social Media ROI
Email Marketing that Works (2nd Edition)