For search engines and search engine marketers, a holy grail has been figuring out how to attract the big money spent on branding into search. Now a new study gives both parties ammunition to say search can build branding.
Produced by Nielsen//NetRatings for the Interactive Advertising Bureau, the survey found brand gains from search ads in a variety of areas.
Not surprisingly, the press release for the study suggests that brand advertisers "may be missing out on a share of voice opportunity" if they fail to buy search ads.
No disagreement from me on that -- or on the idea that search is an important part of any online marketing campaign, something else the press release urges.
How was testing done? For the search ads, there were three groups. Each saw a "generic" search results page that listed ads in numbered order, 1 through 15, under a Sponsored Results heading. No "free" listings were on this page. All listings were inline.
The first group saw an ad with a brand name in the first position, say one mentioning Ford. Another group saw the same ad in the fifth position. A third group didn't see the brand name mentioned at all on the search results page. This last page was the "control" from which any rise was measured.
It's also important to note that the listings were all shown inline at the top of the page under a "Sponsored Listings" heading. For those who have ads in positions that show to the side of results, as often the case with Google, these results almost certainly aren't applicable. Conversely, for those who have top "inline" listings for free, the brand findings in this study might be applicable even to these free ads.
Raising Brand Awareness
One test was of unaided brand awareness. In other words, did the ads help you naturally recognize a brand that was named without being prompted about it. All three groups were asked:
When you think about [industry”, what [companies” come to mind? Please fill in the companies you can think of...
- Control group that saw no brand ad: 43.9 percent still named the target brand
- Group that saw brand ad in fifth position: 49.9 percent named the target brand
- Group that saw brand ad in top position: 55.6 percent named the target brand
You can see there's a pretty nice "brand lift" from the control group figure, where the top position figure is 11.7 points higher. If you were only buying ads to sell things, this was a nice branding side benefit to gain. Not covered in the study is also the fact that anecdotally, others have reported that a brand name in an ad might help clickthrough.
Ads were also divided into six categories: Health, Auto, Beverage, Electronics, Retail and Finance, so that brand lift could be measured in each of those. The lift by category for being in the top position was:
- Health: 19 percentage points higher
- Retail & Financial: 16 points (tied)
- Beverage: 14 points
- Electronics: 14 points
- Auto: 2 points
Ads Or Recommendations?
Interestingly, the results also found that ads weren't viewed as advertising by most of those seeing them. Each group was asked this:
For which of the following [CATEGORY OF COMPANY” have you seen or heard any Internet advertising for recently?
Of the control group, the ones who were NOT shown a search ad for the brand, 8.1 percent said they'd seen an internet ad for the target brand recently. This ad obviously was seen outside the testing.
Now what about the two groups that did see search ads in the testing. If they considered these ads to be "internet adverting," as the question asks, you'd expect to see a big jump from the control group's figure. Instead, only 9.5 percent of those who saw an ad in fifth position and 11 percent of those who saw in in the first position said they'd seen an ad -- a tiny, tiny change in both cases.
This suggests that despite the ads being marked as "sponsored," those viewing perhaps interpreted them as editorial recommendations by the search engine. If that's the case, it's one reason why the brand lift may have happened.
Contextual Not Search
It did disturb me how survey tries to pigeonhole contextual ads as part of search. Search ads and contextual ads were measured separately in the survey, which was great. They are completely different creatures. But the press release slips into making them both fall under the "search" umbrella, which can be confusing to would-be advertisers. Some examples:
- The study was called "Internet Search Brand Effectiveness Research," when "Sponsored Text Effectiveness" would have been a much more accurate umbrella term.
- Part of the release says, "The study demonstrated that sponsored text advertising in the search environment works for an array of branding objectives." The suggestion is that contextual ads are somehow part of a "search environment."
- The study says "Search articles pages generated an aggregate lift of all brand metrics by 15 percent." What's a "search articles page?" That's just a web page with a contextual ad on it. There's nothing "search" about it.
The actual presentation about the survey given as part of the IAB's road show on search also uses terms like "contextual search advertising" and calls contextual ads things like "search text leaderboard," "search text wide skyscraper" and "search text rectangle."
Nielsen//NetRatings senior director of analysis Marc Ryan said that contextual ads were included in the study because some advertisers have assumed they may perform the same as search ads, in terms of having little branding value. The report wanted to spot any differences.
Consider Branding As Side Benefit
Ryan also stressed that the report isn't a call for using search and contextual ads as a replacement for other branding mechanisms. Instead, the findings are intended to help highlight that such ads may have other benefits beyond direct response.
"What we're saying is that there are lots of reason to go into search. By no means are we trying to make the case that you should forgo display ads. You should be combining both," he said. "You also need to measure for everything, to discover additional benefits."
Jupiter Research's Gary Stein stressed the same thing earlier this year in his Dragging Search Above the Line (by the Scruff of Its Neck) article. In it, he shared how nearly half of those in a Jupiter Research survey said they had branding as a search campaign goal. Yet, only 20 percent said they measured any brand gains -- compared to 70 percent who measure clicks.
Why Not Sell On CPM?
Overture did a somewhat similar branding survey back in 2001 that seemingly had the same goal as this recent survey -- attract that branding spend (more about that study is also here: Search Engine Marketing and Branding).
My question at that time was that if indeed search helps branding, then shouldn't it be sold on a cost per impression (CPM) basis rather than on a cost per click one? After all, the brand marketer wants exposure. Recently, I thought this was even more the case in terms of contextual ads, especially when Google now allows contextually-targeted banner and other image ads.
I asked Google, Overture and the IAB for comments about this. Google said because of its quiet period, it couldn't reply. Overture declined to answer and the IAB said any comments on this were up to its members.
Of course, given that branding is only a side benefit, CPM pricing probably doesn't make much sense for the moment.
The study was commissioned for the IAB's Search Engine Committee, a group that includes representatives from major search companies such as Google and Overture and large search marketing companies. This group has an obvious interest in convincing advertisers to spend more on search. Nevertheless, I've heard plenty of anecdotal evidence that search can contribute to branding, and the findings seem to support this.
What do advertisers think? ClickZ has an article with few comments. Kevin Lee also provides a rundown on how the survey was done, the first in a series of articles he's doing looking at the survey (watch for the rest to be posted here). Kevin Ryan also digs into the story here. David Cohen pushes back on the study, warning it's not a replacement for regular online brand advertising. Finally, don't forget to see our recent SearchDay coverage of a conference panel on search and branding: Search Engine Marketing and Branding Challenges.
Want to comment or discuss this story? Come join in at our forums in this thread: Branding Boosted By Paid Ads, Research Says.