Advertiser Wishlist - Combination of Search Query Performance and Geographic Performance Report

A combination of Search Query Performance Report (SQR) and Geographic Performance Report could provide an insight about the type of queries which are being triggered in a Country/Territory, City and Regional view of Keywords triggering impression and clicks. This combined approach could then guide any advertiser to open a separate campaign and target specifically the region and type of keywords.

Unfortunately at the moment we do not get these kind of reports but rather a separate report to check the Search Query Report and Geographical Performance Report. The workaround could be separating the campaigns and targeting them separately but this is not enough to see granular view of query data. The hidden information on how many clicks each of these keywords are being triggered in each region could help advertiser not only target the campaign but also lead towards classifying the marketing 3Ps (Product, Price and Promotion) separately.

If the optimization of keywords are being done with the aid of SQR and we are not considering the origin of the clicks than removing a country/territory, city and region than we are making a biased decision and are rather cutting off keywords considering the entire campaign. This is true for those advertiser whose Campaign are targeting multiple country at a time. Thinking out of the box, one campaign per country targeting is essential but than this would lead towards a massive change over the account and campaign structure. Hopefully, these kinds of services are available in the future, hence my wish list for Search Query Performance (SQR) and Geo Performance Report would hit the top of my request.

Posted at at 10:30 AM on Saturday, November 20, 2010 by Posted by Rabin Gupta | 0 comments Links to this post   | Filed under: , ,

Optimizing Campaign Using Geographic Performance Report

Geographic Performance Reports provides valuable information of the origin of clicks. There are various way optimization over the campaign can be done. Either by optimizing the keywords, adgroups or even by reviewing the CPC bid. These bid optimization could lead towards slashing down average position or reducing traffic by pausing keywords or Adgroups.This could be ideal traditional way of fine tuning campaign. But one vital source information is not considered by various advertiser when it comes to campaign optimization i.e. fine tuning Campaign with the aid of Geographical origin of Click.

Geographic Performance Report is one of the most efficient way of optimizing Campaign. This report provides the actual origin of the Clicks from Country, Region, Metro, and City. This report could be drilled down upto Adgroup level. This report is available in the Google Analytics but for those advertisers not opted into Google Analytics, Google Adwords Geographic Performance Report could be accessed from Client Report at MCC level and inside a Campaign at dimension tab. This report should be utilized efficiently to re-target the location.

For Instance -
An account in language English having various location target would result clicks and impression being generated from various location. It is wise to take a look at the Geographic Performance Report and optimize the campaign such that those Clicks origin from places which are not resulting into conversion  could be removed from the location targetting. These extra dollars spend could then be placed into those places where the campaign is driving conversion. 

Geographic Performance Report does not include the Average position and Quality Score it arguably should not be considered as the most important metrics while reviewing for Origin of Clicks to fine tune campaign. If there is enough clicks and impression present to make decision the average positions and quality score would not be vital as it is out of necessity to cut off those location which are not performing.
Note: If the data is compared from Geographic Performance Report vs Actual Campaign the stats would not be exactly the same. Please note that only placements with negligible number impressions (not clicks) per day may be excluded.

Posted at at 6:43 AM on Monday, November 15, 2010 by Posted by Rabin Gupta | 0 comments Links to this post   | Filed under: ,

Difference Between CPC and Goal/Value Per Click

CPC (Cost per Click), a metrics which revolves around driving visitor to the website and in return accruing costs. For every click there is a cost associated (if  the model revolves around Pay Per Click). Overtime if CPC starts increasing for a keyword  it could be due to poor quality score or there is tense competition in the market. Mostly during peak season the CPC tends to increase as every advertiser would become aggressive.

CPC = Cost / Click

Value Per Click is a metrics which revolves around the ratio between goal or conversion and clicks. This metrics will allow us to measure the quality of the clicks. In theory if the number of clicks increase over the site the conversion should also increase. More clicks should result towards more conversion. If the clicks are increasing and conversion remains stable then the quality of click being bought is not good.

Conversion Rate = Value or Conversion / Click

Increasing CPC would lead towards more clicks and these extra clicks should result into conversion. Quality of clicks should always be placed into radar and this could be done by measuring conversion rate. It is obvious that conversion rate goes down initially as new clicks bought are to be optimized. Once proper optimization over the keywords is done the conversion rate or value per click would become stable. A 2.00% conversion rate is always good.

Posted at at 2:44 AM on Sunday, November 14, 2010 by Posted by Rabin Gupta | 0 comments Links to this post   | Filed under: , , ,

Improving Lost Impression Share Lost due to Budget or Ad Rank

Impression share is a metrics which could be seen at Campaign level. It represents the percentage of time ads were shown out of total impression accrued in the same target market. These loss in impression share could be either due to less budget or due to quality score result rank lost. The lost in impression share means missing opportunity. Sadly impression share metrics is only available at Campaign level, it would be great to see the impression share metrics at adgroup and keyword level. Having impression share metrics at adgroup level would allow us to view the actually opportunity lost for each adgroup.

IS Lost (Budget): The budget for the Campaign is to be set to an amount which does not ends within few hours of the day. This will result not to run the ad as a result of lack of budget. This one is relatively straight forward to fix, increase the budget and let the Ads run all along the day and night. Lost in impression share due to budget for small advertisers could be common. Yes, it is well understood that there is budget restriction for every advertiser. It would be wise to analyze the conversion in the hour of the day and peak traffic hours. Once these hours are noted, it is recommended to run the campaign on scheduled hours.

IS Lost (Rank): This is a tricky one, lost in Rank would be mainly due to lower CPC and lower quality score. This lost in impression share due to AdRank means improving the Quality Score and increasing the bid for the keyword.

How do we Improve Rank/Position?
Rank or Position could be improved by improving quality score and placing CPC above first page CPC. The broad match would trigger tons of query and these query terms are not always being bought. To improve impression share a regular SQR needs to be done

A question would then arise why do we need to do SQR to improve Quality Score? 
There are vaious query terms which comes into an auction these query terms are relevant terms and there could be an intense competition over these query terms. Buying them as exact match would allow us to place different bid and hence appearing in the search results.

For Instance -
My broad match term - buy books online in bangkok  this keyword broad match would also have tendency to capture query like - books online bangkok. The competition for the books online bangkok could be high and the bid placed for the broad term which triggered this query would would lead towards display the ad not in the first page. It might be due to quality score which would make us think if there is need of preparing a new landing page or a new campaign to target these keywords separately.

Hence review over SQR should be done either to include these search terms as keywords or permanently exclude them by adding as a negative exact. This would lead towards improving the ad rank and capturing the lost IS due to rank. Before the work over improving impression share is started, it is a good practice to store the current lost IS due to rank and position, and all other PPC metrics and consider it as a benchmark. And after commencement of lost IS work wait for at least two weeks to measure the results.

Posted at at 7:55 PM on Friday, November 12, 2010 by Posted by Rabin Gupta | 0 comments Links to this post   | Filed under: ,

CPA Model Bidding Strategy

CPA (Cost Per Acquisition) a model revolving around the conversion optimizer. The biggest advantage of this model is that the queries triggered for CPA model are considers origin of users, IP of the user, time the query is being executed, location of the user, browser language setting, historical performance of the campaign etc. Ensure the tracking conversion code is always present in the thank you page. Google Adwords system while opted into CPA model delivers the Ads based on its learning. CPA model is a conversion based optimizer and depends on the Conversion numbers. Although the model is CPA the charge for this model is based on CPC.

It is not recommended to switch CPC to CPA and vice versa. Effect of switching the model usually takes eight  hours to behave normally. It is always good to start the CPA model with slightly higher CPA bid and gradually optimize the CPA bid .While making a switch a default CPA bid is recommended, it is only for a reference purpose and should be updated by the advertiser. The CPA model bid is implemented at Adgroup level. There must be 30 Conversion in a month to have the CPC to CPA model bidding feature automatically enabled. It does get enabled automatically. This would mean at least approximately 7 conversion per week. Moreover it will also depend on total number of adgroups present inside the Campaign. The more conversion before enabling the Campaign to CPA model the merrier it is.

Optimal CPA Bid - Last two weeks Cost Per Acquisition should be used to benchmark the data before applying the CPA bid. Ideally, the CPA bid should be around +/-10% of the real CPA. Revisiting CPA on a weekly or twice per week to ensure the CPA bid is still in acceptable range. Always determine the ROI which you are willing to achieve and  calculate the CPA which you are willing to pay.

  • CPA Bid = Total margin per conversion * ROI  you are willing to achieve
  • Average Cost Per Acquisition (CPA) = Average Cost per Click / Conversion Rate
    Remember, seasonality would lead towards price fluctuation and site performance plays a very important role and hence a constant update would be needed while calculating the CPA. Using a factor for adjusting CPA could be done based on learning. At least two weeks are needed to measure the true performance of the CPA model. CPA model is an ideal model and this could be seen as a future for various advertiser. There is no keywords based bidding required but reviewing SQR to improve the Campaign performance. For small advertisers this could be a blessing as they could opt into CPA Campaign and Google would ensure there CPA goal are met without having to worry about the bid for each keyword.

    Posted at at 7:02 AM on Tuesday, November 9, 2010 by Posted by Rabin Gupta | 0 comments Links to this post   | Filed under: , ,

    Search Funnel Redfines Keyword Strategy

    Search funnel is an insight over the consumer clicks to conversion cycle. Every consumer landing on the page for the first time is unlikely to purchase or commit transaction. Every buying process cycle goes through a phase of research, compare, evaluate or decision making and purchase. These research to purchase phase are vital for every business and for every advertiser it is a must to be present during the entire cycle of research to purchase.

    Every advertisers would take a look at Search Query Report and start optimizing its Campaign, Adgroup such that the keywords being triggered are well targeted. Now, in the process of improving the performance of the Adgroup as a rule of thumb we tend to place search query terms as negative. Let us try visualizing this as an example:

    Supposing, the advertiser is in the business of selling air tickets to be more specific selling air tickets in Bangkok. Structure of the Campaign are as below.

    Campaign Name: 
    • Thailand Air Ticket
    Adgroup Name: 
    • Bangkok Air Ticket
    • bangkok air ticket
    • air ticket bangkok
    • book air ticket bangkok
    • online ticket bangkok
    • purchase bangkok ticket
    • reserve bangkok ticket
    Now, looking at above examples these broad match have tendency to trigger the terms bangkok, air ticket and so on. Ancient approach would be to kill all those keywords which are too generic and are not performing at all. So my first thought would be to add the term bangkok as exact negative. The question should be am i doing right? A traditional answer would yes, correct this is the right approach. If viewed from Search Funnel eyes we should redefine our way of adding negatives or classifying the keyword strategy.Those keywords which are not performing could be assisting keywords and are vital during the purchase cycle.

    Keywords Relationship - Search Funnel - Purchase Cycle

    There is relationship between keywords and these relationship are made more clear if we view it via search funnel. In every stage of purchase cycle generic terms, head terms and long tail terms plays a vita role. Before we proceed on permanently cutting off traffic we should make a strategic decision. If these terms are generic terms or head terms or long tail terms we might want to re-categorize our campaign and adgroup to prepare a proper adtext to target our users.

    Overall Search funnel redefines the way we think about keywords, our bidding strategy and negative keywords strategy. There are keywords which help assist the purchase and killing those keywords could impact the performance of the adgroup. Hence search funnel now could give a clear picture on which assisting keywords are needed an extra optimization to improve the performance of the adgroup.

    Posted at at 3:48 AM on Sunday, November 7, 2010 by Posted by Rabin Gupta | 1 comments Links to this post   | Filed under: , , ,