<< by on June 5th, 2012
Session one of the paid search track, opens with the “Mad Scientists of Paid Search”. ‘Staring’ Michael Behrens of WebMetro, Benny Blum of eSearchVision, Ryan Hutchings of Vacation Roost and Matthew Mierzejewski of RKG.
It’s kind of a “free form” session of assorted ways to optimize PPC Campaigns.
First up is Michael Behrens. Secrets of Google’s Matching Algorithms. Users type queries and keywords are selected. Google doesn’t always match the right keywords with the right keywords, intents and adgroups, etc. After 2008 AdRank became a factor in determining keyword matches.
They ran diagnostic rules to see how often Google Mapped keyword queries to the same keyword. Google “multi-mapped queries” to multiple different keywords. Some keyword queries showed up in multiple adgroups and campaigns. 10% of what you spend on average have “multi-mapped” search queries that show for different keywords and adgroups. Its possible for a keyword to broad match to up to 8 different keywords.
Ad Rank is the likely factor that keeps keywords “hopping” accross to multiple keywords and adgroups.
Only 53% of the keyword matches they tested were “correct” Length of Match Type seems to be the most ignored actual exact matching of keywords. However the Click Through Rates are extremely high on “Match Type Length Queries”.
Use of Negative Keywords are the best way to combat this. You can force correct match typing this way. Account structure with tightly themed Ad-Groups helps. Choosing match types carefully and limiting use of Broad and Phrase match types.
Ryan Hutchings from Vacation Roost is next. Vacation Roost studies the same exact thing that Michael Behrens did. Vacation Roost is a wholesaler of vacation rentals and packages. They broke campaigns into hundereds of adgoups and saw search queries continuously coming through in the wrong ad groups. They spent months breaking adgroups at a super granular level – even broke them out with match type only ad groups. Doing this really increased click through rate and lowered Cost Per Click. The clicks went up substantially too – up 376%. It didn’t solve the problem of the wrong ad being served up.
Google was favoring broad match version over the exact match types. Ad Rank was trumping exact matches, most resrictive match types and cheaper keywords with higher Quality Scores. It favored keywords with the higher ad rank (consequently higher costs and CPCs too). The solution was to add 10′s of thousands of negative keywords to “force Adwords to respect keywords and override the misinterpretation.
“Google has buckets for industries. The system operates within these buckets and will interpret on your behalf within the buckets.” In Google’s eyes a Hotel is the same as a condo, townhome, villa, timeshare, bed and breakfast, inn or “places to stay”. The interpretations trump tightly knit account structure.
Check your ad diagnostics tool to verify Google is showing the right ad for the right keyword.
Matthew Mierzejewski from RKG. He has 4 topics. Bottom Ads, Google Quality Score and Coversion Rates, Bing Rich Ads in Search, the Decline of Google’s CPCs.
Bottom ads are literally at the bottom of the page. Google hasn’t really revealed why they are using them. Bottom ads traffic is low. They have performed better on Click Through Rates than side ads. RKG data says that conversion rate isn’t any better than side ads. RKG thinks side ads have higher Click Through Rate.
Quality Score impact on Conversion Rate. In most cases there is a weak positive association between Quality Score and Conversion Rate. Performance varies by Verticals. Its one of man metrics that can inform bidding.
Bing Rich Ads in Search. They have site links, Video Thumb Nails and Images. CTRs are a double digit increase. Its hard to get them to show and you have to raise bids to get them to show. Bing suggest raising bids to get them to show. Bing doesn’t allow competitors to show when a Rich Ad in Search show.
The “Root of Google’s CPC Decline” RKG says the popular argument is mobile traffic. Smart Phone’s in some cases have 50% lower bids. Brand Queries and Prouduct Listing Ads are the largest contributing factor in the CPC decline. Impressions on Product Listing Ads have really gone up and the costs are much lower for them. Google’s elimination of free shopping listing and switch to Product Listing Ads may raise competition.
Brand Queries are now eliminating competing links and the SERP result pages. 51% increase in brand query clicks than 2 years ago. Lower competition on brand queries have impacted this. Quality Score increases on those Brand Queries have lowered the costs as well.
Benny Blum from eSearchVision. He’s discussing user intent and device behaviors.
There are different search behaviors on PCs, Tablets and Phones. Tablets have higher user engagement. Three types of Advertisers. #1 Impulse: Low AOV. #2 Service or Software – online service offerings. #3 Direct Response – Ecommerce.
Mobile users have extra ordinarily high bounce rates. Mobile time on site is skewed by bounce rates. Time on site is higher on Time on site. Mobile drives heavy new visitors though it has the highest likelihood of new visits. Mobile works well with cheap products.
Purchase intent: Mobile has high click through rate and low cost per click in the service and software vertical. Tablets have low click through rate and low cost per click, but higher relative cost per click than desktop search in the service and software vertical.
Actual conversion intent, had minimal variance in device. Low cost impulse purchase conversion rates are pretty high on Tablets though.