A discussion draft of the Online Sales Tax Simplification Act of 2016 is expected to be introduced in Congress soon.
(Page 2 of 4)
As long as Google is limited to serving ads on its own pages, its revenue can grow only incrementally. But a contextual ad program breaks through that ceiling. “You now have available not just page impressions on your own interface, but all of the page impressions of any web site on the Internet that will carry your ad,” says Fredrick Marckini, CEO of search engine marketing company iProspect.com Inc.
That’s part of the thinking behind G-mail. Introduced this spring, G-mail applies the same contextual ad technology to e-mail, serving up paid ads relevant to the content of subscribers’ e-mail exchanges. Some critics have raised privacy concerns about G-mail, but Fischer points out that it’s technology, not humans, that scan subscribers’ e-mail for content and that no record is kept of that content. Furthermore, G-mail packs its own spam filter. Should any pesky unsolicited e-mail slip by the filter into an in-box, G-mail subscribers needn’t worry about it triggering ads for, say, Viagra. “We are not showing ads for the type of content people might not want to see,” Fischer says.
G-mail has been in test only since April and its success as an advertising tool remains to be seen.
Google also isn’t releasing any results for Froogle, though it’s a mark of confidence that in April it bumped up Froogle from a hard-to-find location in the Advanced Search function to a featured link on the Google home page. If Froogle’s index is basically a subset of Google’s broader one, and a merchant’s pages already are in Google, chances are they’re already in Froogle as well. That said, why should retailers in that position bother with the direct feed to Froogle, even if the feed is free?
The answer is that natural search results on Froogle and Google will likely differ. While Google’s search results for products will draw from content including product reviews or discussions and manufacturers’ pages, Froogle results will show only e-commerce pages. Furthermore, Froogle’s ranking formulas for both natural search and paid listings differs slightly from Google’s, says Froogle’s director of engineering Craig Nevill Manning.
“The standard Google ranking didn’t work as well as we liked for Froogle, so we modified it somewhat,” says Manning. “Then we mixed in some secret sauce from the Froogle side, which takes advantage of the fact that we have structured data such as price and images.”
Froogle gets those data most efficiently from the direct feed, though it also crawls merchants’ web sites regularly in the same way Google crawls sites. “We want everybody to be included in the index, so if a merchant doesn’t give us a feed, we’ll do our best to crawl the site and understand how it’s formatted in order to recognize prices, images and names of products. But we may not always get it right,” Manning says. “And we may not crawl your site as frequently as you’d like for updating prices. The feed lets merchants upload a text file every day with updated information.”
Manning offers a hint to merchants seeking to optimize pages for higher rankings in natural search on Froogle. “When merchants give us a product name, the more standardized that name is, the better,” he says. Under that rule, merchants selling a Nikon CoolPix 5700 digital camera, for example, should list it in exactly those terms, rather than using a longer description such as, “ Nikon CoolPix 5700, great digital camera, and five megpixels with three flashcards.” Says Manning, “That will be more likely to confuse our algorithms in terms of ranking and figuring out whether the item is relevant to a query. People tend to put in fairly short search queries.”
From 100 to thousands
Marketers looking to drive the most gain out of Google’s growing complexity need an expanded strategy. Larkin notes that Performics has in the past six months seen an uptick in new clients who’d previously been running both Google and Overture campaigns on their own. “It used to be they could buy 100 keywords and have an effective marketing program. Now they realize they need to buy thousands if not tens of thousands, so they need help to scale those programs,” he says.
And help abounds for those willing to pay search engine marketing firms, which have developed different approaches to Google. The first advice Marckini gives clients attempting to figure out what advertising to buy on Google is to not buy anything before optimizing pages to Google. “Optimize your site for natural search first because in Google, that’s where 70% of the clicks are happening,” he says.
As a next step, marketers should consider buying Google’s paid AdWords. “30% of the people who are on Google are clicking there-and they may be a different 30% who are clicking on natural search results,” Marckini adds. The next step is the direct feed to Froogle. While not the most heavily used shopping engine, he says, it has enough traffic to have an impact on revenue. Then, depending on their market, retailers should look at Google’s contextual ad program, AdSense.
“There are reports contextual advertising doesn’t convert as well as paid search. But our experience is that it introduces you earlier in the buying cycle to the consumer who’s doing research,” Marckini says. “The customer may not buy then, but they may go back and buy it on your search ad later. The key is paying hyper-close attention to the conversion rate on these programs, because you can spend an awful lot on keywords that don’t convert at all.”
That’s why Lisa Wehr, president of search marketing company OneUpWeb, advises clients looking for a new Google strategy to start by closely examining what they’ve already done. “If they’re a current AdWords advertiser, they should look at what are the keywords and conversion ratios they’ve experienced,” Wehr says. “One of the biggest holes I see is that marketers aren’t tracking this, so they could be flying blind.” While Google does offer a conversion tracker, OneUpWeb expects to roll out an independent tracking tool next month.