Live blog coverage (@bizwatchlaura) – please note that I will be adding only the most useful information related to paid search during/after this presentation. If you attended this session and want to add any of your comments, please let me know by commenting below.
Andy @pronouncedAhndy is talking about close variants (pen does not equal pencil). Exact (close variant) non-brand conversion rate relative to pure exact match median advertiser. He is seeing a slight negative impact on conversion rate in Q2 2017 on non-brand conversion rates. One possible scenario if you have target CPA bid strategy and you are dropping slightly in performance (ROAS), you could have keywords get turned off if that performance drops even by 1-5%. However, he is not yet seeing considerable negative traffic impacts. Note: We are seeing the same. Bizresearch is not seeing negative impacts from close variants on exact match.
Ad Rank Changes @AdWords – Minimum ad ranks topic. To be eligible for a top spot, your Ad rank needs to meet a minimum threshold. As a result, CPC when you appear above search results is generally greater than the medium bid. (google documentation – will look for URL later). First page minimum vs top of page minimum. He is seeing substantial increase in non-brand in those first page & top of page bid requirements.
Dr. Laura spoke from @microsoft. Has an undergrad in biology, grad in microbiology. She works in planning & forecasting at Bing. Research can’t isolate one reason for improved CTR on image ads in the original marketplace. It could have been due to higher rankings, and not related to image ads. Today’s ad & data analysis marketplace might not be search queries that are really related to commercial queries. For example, UK children using Bing might show increased search queries around a particular holiday, which may not have anything to do with searcher intent, or commercial queries. Multiple experiments are required to fully understand the data.
Hypothesis 1: Very granular account structure harms performance. Google recommends that you separate out your Google Shopping product ads (after 5 clicks), that it would be better. But he is not seeing any improved performance when products are broken out into separate product groups.
Hypothesis: Most interesting is that cheaper products get better impressions. The more expensive products get shown less.
Hypothesis: Product price determines ad’s position. Two types of product position. Offer position vs Product position. The offer position (data provided by price comparison tool) with the product cheapest price usually shown highest, except in one case where no seller rating might have adversely affected visibility. Another situation where lower position despite better price – they are bidding one price that is .50 lower than Google’s recommended bid. They did a test on expensive products vs cheap products. Despite bids, expensive products just did not get the same impressions as the cheaper products.
Do you invest in high CPCs or afford a price reduction?
Hypothesis: Do you manipulate product titles for better performance? or description? Nothing really happened when the product description was purposefully done to inaccurately describe the product. What did matter, spend time on your product titles.
Hypothesis: Changing titles led to historical data loss.
Performance only saw a slight dip on the day of the feed upload. Hypothesis proved false.
Hypothesis: ECPC uses audience information to predict conversion probability in Google
Does it overwrite RLSA modifiers? They started switching on the ECPC (enhanced CPC). Google pushed lower funnel audiences due to higher conversion probability.
Hypothesis: Does raising bids in shopping give you bad traffic? Do you get less relevant search queries/users, clicks?
Tested design: increased bids on brands by 200%. Raised bid from .50 to 1.50. Did get more impressions but conversion did not improve or stay the same. Using query length as indicator, what happens to traffic quality when increasing bids. Raising bids attracts a higher share of shorter search queries, a similar observation as with broads. Long tail inquiries increase at far lower growth rates.
- Are searches / click data coming from within Google Shopping or Google SERPs where text ads appear alongside shopping ads. Andreas thinks roughly 85% might be coming from SERPs, remaining coming from Shopping. Those who use Google Shopping might be more price sensitive shoppers/searchers.