Bid adjustments are besides the keyword management the most important way to optimize the performance of your campaigns. In this article you will learn how BidX handles this part with our powerful Automation for you.
Important: First, learn how to create an automation and read numerous answers about what exactly the automation does.
The basis for an adjustment are the aggregated performance data since the keyword bid was adjusted the last time. Therefore BidX considers its own adjustments, those from other tools as well as manual bid adjustments. The performance of the last 72h is not taken into account, since this time is needed for the attribution of sales. Besides keyword performance data also search term performance data, ad group performance data and campaign performance data are considered.
With having ad group and campaign performance on hand it is possible to calculate their ACOS. Keywords for which no ACOS exists can refer to the next higher performance hierarchy of ad groups or campaigns. This also applies for the conversion rate. Bids are increased but only if the current ACOS of the ad group and campaign is below the ACOS threshold. However, bids are only increased up to the max. bid which consequences from the following formula:
max. CPC = Selling Price x TargetACOS x Conversion Rate
If no sales are generated even after a specific amount of clicks, bids are reduced tremendously and keywords respectively search terms are set as corresponding negative keywords but always also as positive keyword added.
To make sure the conversion is not a random event, a specific amount of clicks has to be reached. If an ACOS is much higher than the target ACOS, the bid will be lowered accordingly and the other way around. If the ACOS is much higher than the target ACOS the keyword or search term will additionally set as corresponding negative keyword. However bids are only increased or decreased by a certain percentage of the current bid.
Generally the main question is how to find relevant keywords that will most likely convert in future and adjust those that already converted accordingly. In principle, the target is to maximize sales under the constraints of an upper ACOS limit or monthly budget.
Another question is how to recognize keywords that will not convert in future and cut their bids early to guarantee that half the money which is spent is not wasted.
After detailed data analysis using machine learning five steps for the optimization can be favored. Basis were 300M data sets whereas one data set represents the daily performance for a search term including impressions, clicks, cost, conversions and sales. All other KPIs can be calculated on this foundation.
1. Phase: Optimize for Relevance
For the ad to be displayed some requirements have to be fulfilled. Not being displayed for certain keywords is one of the main problems sellers and vendors face. To point out the most relevant points for this:
- Sponsored Products can only be shown when a seller is the owner of the Buy Box. For vendors Amazon Retail has to be the owner. Consequently there is no competition during the auction within a group of sellers who are selling the same product. As well you will not bid against yourself unless you use the advertising console and Seller Central at the same time for the same keywords.
- If a search term offers only a few results, Amazon might not show any Sponsored Product Ads. Hence, the search volume for a search term has to be efficient.
- The bid has to be high enough and competitive. If the bid is the 5th highest, but only four ads are displayed, an advertiser has to increase its bid to be displayed.
- Last but not least ads appear only for relevant search terms. Amazon’s A9 evaluates based on historic organic performance if an ad is served for a specific keyword. Important is that the product is in the right category and it is ranking organically for the search term. This can be verified by entering the ASIN and the specific search term into the search field on Amazon’s website.
Keyword bids will not be increased automatically in this phase at the moment. It can be realized with an individual rule.
2. Phase: Optimize for Click-Through Rate
According to Amazon the CTR scoring is one of the most important relevance factors that impact if an ad is served. Advertisements with an extremely low CTR will be cut off. A search engine result page (SERP) slot consists of the following information which can influence a buyer’s decision:
- Main picture
- Title
- Price
- Plus-Product
- Review amount and rating
- Best-Seller Badge
- Date of delivery
- Prime eligibility
- Vouchers
Most of the information can be split tested to gain the best and significant results. Tests can be either realized directly with Amazon ads or with third party services.
When analyzing the influence of CTR on keywords for certain bids the data states that if the CTR is below a certain threshold a sale is unlikely to happen. In this case the keyword bid will be decreased. On the other hand, with a CTR higher than a certain threshold a sale is likely to happen and the bid will be increased.
The system will wait until it makes an adjustment. Read in our blog how many clicks it will wait approximately: Economic Clicks
3. Phase: Optimize for Conversion Rate
As soon as the CTR crosses the threshold the bid will be increased. This results from data where keywords with higher bids have a higher possibility to convert. This is reasonable since the keyword will be displayed more prominent and has a higher visibility and SERP information seem to suit the customer.
4. Phase: Optimize for Advertising Cost of Sale (ACOS)
As soon as a keyword is converted it is feasible to calculate the ACOS. This is the first time where it is possible to adjust the bid based on whether the ACOS is above or below the target ACOS which was set in the beginning. Bids will be adjusted dynamically based on the distance between the target and the actual bid. A suitable solution that we found is to adjust the bid depending on it's distance to the target. For instance, if the target ACOS is 25% and the current ACOS is 35% the distance is 10% and the bid will be decreased accordingly - same counts for increasing. Advertisers have to make sure that the target ACOS is set realistically. Too high adjustments in a single step may lead to revenue drops because one of the more competitive ads is not delivered on 1st page anymore, but 2nd or 3rd page. The bid adjustment is therefore limited for a single step up or down based on the current bid.
5. Phase: Optimize for Profitability
For this, it is important to look at the big picture. This includes revenues generated organically and through the use of advertising subtracted by all product costs, general costs and advertising costs. As a result, sometimes a higher ACOS can generate a higher overall profit, assuming the higher ACOS is below the break-even-ACOS. However organic sales data have to be taken into account which is not the case at the moment.
If you like to learn more on how the keyword management works, read more over here: How does the keyword management work in BidX?