Targeting Search Ads By Demographics & Behavior

Dec 5, 2005 • 3:30 pm | comments (0) by twitter Google+ | Filed Under Search Engine Strategies 2005 Chicago
 

First up, Jed Nahum from MSN Search and he explains that targeting is what it is all about. He went over some of the highlights of adCenter, they piloted in France and Singapore earlier in 2005 and now they have a US pilot. Audience intelligence drives ROI. Learn about your customers and connect via rich targeting and then refine your campaign. Learn about your customers, he shows a slide about a keyword research tool that has all these cool demographic tools within adCenter. He compares Oprah versus Dave (letterman). Oprah has more female centric, Dave is more male centric for searches. Dave's is a bit younger and Oprah's is older. Dave's audience is a bit richer. But in fact, Oprah has 20x more searches then Dave. That is why Oprah, herself is a bit richer then Dave. :) Connect through targeting (bid boosts); demographics, geographic, day of the week and time of day. He then pulled some comments from customers on how cool the tools were. Refine Your Campaign; he shows a rich report with lots of data.

Next Up Roy Shkedi from AlmondNet. 40% of internet ads dollars are spent on search engines where people spend less than 5% of their online time. On the majority of the sites where people spend the other 95% of their online time - the ad supported content is sold for very low CPMs. After people search, Almondnet presents people with additional paid ads. So basically you do a search in Google, lets say on "health insurance", then later, he is on fox search, the ad space on fox news is then served up to the banner space on fox news. It basically tracks your searches and then serves up ads on other non-search sites through contextually relevant banner ads. Advertisers can now serve up ads after the search, there is less competition on these pages versus the search engine pages, free branding, reach consumers at multiple stages of the purchasing cycle, clicks originating from behaviorally target ads convert much higher. The ads must be targeted based on a recent purchase intent versus what a person happened to read.

Kevin Lee from Did-It now up, he is really tall :). He shows the heat map of how people look at SERPs, golden triangle. Why targeting matters? Marketers get to put more of their budgets towards their best customers, searchers only see the ads that the marketers really want seen, and publishers get a higher yield on their search and impression inventory. The key is to find out what your power segments are. MSN adds increased targeting options by allowing you to raise your bid by demographics (age, gender, geographic), the additional targeting allows you to be more effective. Same deal with day parting. Behavior Segmentation; better conversion from click to lead/sale, higher immediate value, better lifetime value of the customer, offer responsiveness (look at your data and determine if behavioral targeting makes sense). The engines are also targeting text and graphical ads behaviorally based on prior search behavior, prior click behavior and content preferences. Its a lot of trafficking work to restructure campaigns to market differently to segments. Leads to better conversion from click to lead/sale. Higher bids may get you additional volume. Higher lifetime value... Offer responsiveness, does one segment respond to a different offer? The perfect storm, happens when the different segments overlap (you can bid much higher). Choosing demographic - behaviors; look at your immediate conversions rates by geo or time, do an analysis of your current customers as they convert. If MSN or your conversion data indicates a power segment uses certain keyword phrases, you can media buy against the keyword; which sites show up high organically, what contextual network do they belong to? do they sell ads directly. You may be able to get the visitor one click after the SERP.

Danielle Leitch from More Visibility next up. Do you know who your customer base is? Is your customer base very niche or learning towards a particular subset group within the categories above? Why not do this in search; you are probably doing it offline now. PPC Evolution; chose keywords, bid prices and added ad copy. Since then there are many new options; bid by ROI, demo, behavioral, by engine, local, day parting, contextual, and advanced tracking. Research just gets better daily, incredible enhancements can be made to your keyword analysis and selection process. She also shows off MSN adCenter screen captures. Make bid boost decisions, aid in keyword research selection, start using it to identify areas of campaign expansion or optimization. Geography profiling is very exciting, shows off MSN adCenter again. She then tells us to expect more from Yahoo and Google and see more enhancements in targeting. She said its not all 100% accurate, because hotmail info may be fake. But behavioral search will continue to improve.

Dana Todd from SiteLab, not too tall, but does had really red hair. She said that its funny this is not called adware and spyware buys, since this is what it was called a year ago. All of a sudden, now its legit. She talks about Yahoo! Fusion Marketing. Targets consumers with specific affinities & interests derived from online behavior (terms searched, pages visited, ads clicked and product bought). Display behavior 3 or more times in a 28 day period is how Yahoo's product works. You can buy "channels". Impulse targeting reacts towards search. She brings up a case study, hot spring spa; they were buying keyword targeted banners in search in the old days. Performance was dropping. they decided to give behavioral units a try. They ran monsters, LRECs and other units. She shows a graph of the decline and improvement later on.

One thing, seems like all the people on the panel love the data from MSN adCenter. FYI - I believe Chris and I overlapped, so we only covered one session this track...sorry.

Previous story: MSN Contextual Ad Program Coming Soon?
 

Comments:

No comments.

blog comments powered by Disqus