Automating "how much = bid" based on past results
Posted: Sat Dec 21, 2024 3:55 am
AdWords has several purposes for using automatic bidding (such as ad position and number of clicks), but here we will focus on the bidding type that uses "number of conversions and revenue" as an indicator.
There are three bidding types available that use "number of conversions or revenue" as an indicator.
Extended CPC
You set a cost-per-click bid, and your bid will automatically be adjusted within a range of -100% to +30% of that amount depending on the likelihood that a click on your ad will lead to a conversion.
Conversion Optimizer / Target CPA
Your bids will be automatically adjusted to maximise the number of conversions within your target CPA range.
Target Return on Advertising Spend (ROAS)
It predicts the probability and value of conversions and automatically sets bids to maximise the value of those conversions, whilst striving to reach a target average return on ad spend.
This time, we will look at the performance phone number database india of automated bidding by comparing the results of using "Conversion Optimizer / Target CPA" (hereinafter referred to as CO) with those of manual bidding.
Goals are set according to the number of sales personnel allocated to each area, so designs are made on an area basis.
CO was introduced carefully to this account, and was introduced in stages to limited campaigns. Therefore, manual operation continued in the Tokyo, Chiba, and Saitama areas, and CO was introduced only in the Kanagawa area. The above is the performance 45 days after introducing CO, and the delivery results for 10 days. As you can see, the campaign with CO applied achieved much higher results than the manual bidding campaign.
Period from CO introduction to screen capture: 30 days
04-2
For this account, we decided to gradually introduce CO after checking the results. Although the CV and CPA were unstable for about two weeks immediately after the introduction, the results were close to the target values after that, and the number of cases has been steadily increasing.
Because campaigns are designed for each product category, the cost per customer and profit margins vary for each campaign, but the target CPA was set at a flat 10,000 to 15,000 yen after calculating the overall average cost per customer and the repeat rate per period, and the KPIs were simplified for operation.As a result of introducing CO, it was possible to lower the CPA while maintaining the number of cases.
Since then, both of these accounts have implemented CO for 70-80% of their campaigns, automating a large portion of their bidding processes .
*Some may say that information showing performance with screen captures like this is lacking in evidence, as it often does not mention the process that led to the results or external factors. However, even if you ignore that, you can see how practical it is, as it returns this level of performance when you are "fully entrusting the bidding process to them."
There are three bidding types available that use "number of conversions or revenue" as an indicator.
Extended CPC
You set a cost-per-click bid, and your bid will automatically be adjusted within a range of -100% to +30% of that amount depending on the likelihood that a click on your ad will lead to a conversion.
Conversion Optimizer / Target CPA
Your bids will be automatically adjusted to maximise the number of conversions within your target CPA range.
Target Return on Advertising Spend (ROAS)
It predicts the probability and value of conversions and automatically sets bids to maximise the value of those conversions, whilst striving to reach a target average return on ad spend.
This time, we will look at the performance phone number database india of automated bidding by comparing the results of using "Conversion Optimizer / Target CPA" (hereinafter referred to as CO) with those of manual bidding.
Goals are set according to the number of sales personnel allocated to each area, so designs are made on an area basis.
CO was introduced carefully to this account, and was introduced in stages to limited campaigns. Therefore, manual operation continued in the Tokyo, Chiba, and Saitama areas, and CO was introduced only in the Kanagawa area. The above is the performance 45 days after introducing CO, and the delivery results for 10 days. As you can see, the campaign with CO applied achieved much higher results than the manual bidding campaign.
Period from CO introduction to screen capture: 30 days
04-2
For this account, we decided to gradually introduce CO after checking the results. Although the CV and CPA were unstable for about two weeks immediately after the introduction, the results were close to the target values after that, and the number of cases has been steadily increasing.
Because campaigns are designed for each product category, the cost per customer and profit margins vary for each campaign, but the target CPA was set at a flat 10,000 to 15,000 yen after calculating the overall average cost per customer and the repeat rate per period, and the KPIs were simplified for operation.As a result of introducing CO, it was possible to lower the CPA while maintaining the number of cases.
Since then, both of these accounts have implemented CO for 70-80% of their campaigns, automating a large portion of their bidding processes .
*Some may say that information showing performance with screen captures like this is lacking in evidence, as it often does not mention the process that led to the results or external factors. However, even if you ignore that, you can see how practical it is, as it returns this level of performance when you are "fully entrusting the bidding process to them."