Apply a Bid (Dynamic) Multiplier vs. Fixed Bid Adjustment

Quick Links

Introduction
Why You Should Apply a Bid Adjustment
Risks and Best Practices
Bid (Dynamic) Multiplier vs Fixed Bid Adjustment
When to Apply Bid (Dynamic) Multiplier vs Fixed Bid Adjustment
FAQ's

Introduction

Bid Adjustments is a powerful feature that integrates Human + engine interaction to strategically experiment with Perpetua's Ad Engine. This feature aims to incorporate human logic into the engine's bidding algorithm to enhance its performance.
The Perpetua engine will consider the data collected from the bid adjustments referred to as experiments to train the algorithms for your specific account, incorporating both positive and negative results.

You can choose a fixed bid or a bid multiplier adjustment which will be optimized for a set duration that you choose.

🚨 Note 🚨
Bid Adjustment applied to Custom Goal or Mutli-Ad groups taken over campaigns will not be shown in the "Experiments" tab.

Why You Should Apply a Bid Adjustment

Drive spend on keywords with no impressions but are deemed strategic.
Reduce spend on poorly performing or “expensive” keywords that are not part of your strategy.
Drive spend on well-performing keywords.
Experiment with new keywords the Ad Engine is not spending on.

Risks and Best Practices

If a bid adjustment is set up inefficiently, whether due to wrong keywords or using too many at once, it can increase the realized ACOS of your goal. Since the adjustment is a manual override on the bid, the ad engine may need to reduce bids on other targets to stabilize your ACOS. This could potentially lead to a decrease in performance.

To prevent this scenario, here are some best practice recommendations:

Apply bid adjustments to a maximum of 3-5 keywords in a goal at one time.
Always set an end date on your bid adjustments instead of leaving them "Always On". However, if this aligns with your strategy, consider launching a Keyword Boost Instead learn more here).
Aim for a duration of 7-14 days depending on the aggressiveness of the change. This allows the engine enough time to gather sufficient data and for you to observe results.
Ensure your bid multiplier is high enough to make a noticeable difference in performance but low enough for you to tolerate if performance is inefficient, minimizing risk.

Bid (Dynamic) Multiplier vs Fixed Bid Adjustment

Bid Adjustment Type	Description	Engine Interaction
Bid (Dynamic) Multiplier	Use a bid multiplier if you want a dynamic bid that will change over time.	Perpetua will multiply the bid it calculates by the value you input.
Fixed Bid	Use a fixed bid if you want to choose the exact bid price.	The bid will be permanent for the chosen duration and Perpetua will not optimize/change it.

🚨 Note 🚨

Setting the Bid (Dynamic) Multiplier Adjustment to 100% will keep the bid value unchanged.

⬆️ Upward Bid-Multiplier Adjustment: any value above 100%
⬇️ Downward Bid-Multiplier Adjustment: any value below 100%

When to Apply Bid (Dynamic) Multiplier vs. Fixed Bid Adjustment

📗 Bid (Dynamic) Multiplier

Scenario	Considerations	Suggested Action
Keyword ACOS <= than Goal/Segment Target ACOS	The Perpetua ad engine is already increasing the bids on the well-performing targets driving sales by referring to the Target ACOS you applied to the goal(learn more here). Applying a bid multiplier adjustment can further enhance results.	Upward Bid Adjustment Consider applying a bid multiplier to targets with a Realised ACOS <= than your Goal/Segment Target ACOS with a value higher than your keyword CPC. This will accelerate the performance of the target. Maintain this multiplier for approximately 14 days to allow the engine to experiment with the new setting.
Reduce Spend on Poor-Performing Keywords	While you can achieve this by pausing targets, lowering your bids with a downward bid multiplier offers more flexibility. This approach allows you to set a duration and control the specific bid for that target, rather than stopping bids entirely.	Downward Bid Adjustment Identify targets that are spending without generating any sales. For example, you can apply a filter to your target list using the following criteria: Filter by: Performance Clicks >=10 Attributed Sales = 0 Apply a downward bid adjustment to reduce spending. Keep in mind that Perpetua already performs this action (learn more here), but a bid adjustment accelerates the process.
Experiment with New Keywords	If you're introducing a new set of keywords that you've never bid on before and you want to assess their performance over a specific duration, it could involve advertising seasonal products you're interested in experimenting with, or keywords the engine is not spending money on)	You can choose between using either a fixed bid or a multiplier approach. To determine the appropriate value to apply, refer to the average CPC over the past 30 days within the goal

📒Fixed Bid

Scenario	Considerations	Suggested Action
Keyword ACOS > than Goal/Segment Target ACOS	Applying a bid multiplier might still cause the engine to lower the bids on these targets to achieve your Target ACOS. (learn more here).	Set fixed bid values higher than the target's CPC for those targets. This will increase the likelihood of your target winning the Amazon auction. Ex: CPC = $0.85 Fixed Bid Suggested = $1.20 Active for 14-30 days.
Advertising Products With Little Advertising History	Without (or low) advertising history on your products, the Perpetua ad engine might face difficulties in placing competitive bids	A fixed bid directs the engine to bid the specific amount you set, allowing it to learn from its performance over the period the bid is in place. Ex: CPC = $0.85 Fixed Bid Suggested = $1.20 Active for 14-30 days.
Drive Spend on Strategic Keywords	You may find that keywords relevant to your products are not performing well. Why? Perpetua's engine optimizes your keywords based on the sales they generate. Even if a keyword is closely related to your advertised product, the engine's bid optimization prioritizes the keywords that drive sales. The engine will still apply experimentation according to the Target ACS you apply, but if the keyword does not generate sales, the bid will be lowered (learn more here).	Applying a manual fixed bid to those keywords helps guide the engine to increase competitiveness and improve the chances of winning the Amazon auction. Set a fixed bid that is relatively higher than the current CPC for that keyword. Ex: CPC = $1.25 Fixed Bid Suggested = $1.40 Active for 14-30 days.
Experiment with New Keywords	If you're introducing a new set of keywords that you've never bid on before and you want to assess their performance over a specific duration, it could involve advertising seasonal products you're interested in experimenting with, or keywords the engine is not spending money on)	Without historical data, using a fixed bid for these targets ensures more consistent spending based on the set bid amount. To find the appropriate value, refer to the average CPC from the past 30 days within the goal.

FAQ's

Click on the ▶️ to view our considerations and suggested actions.

Why did Perpetua apply a lowered bid compared to my historical ones?

Considerations

Perpetua gets access to target history, by harvesting targets that were well-performing during your 60 days before launching a goal on Perpetua, however, we do not have access to the bid’s history.

For instance, if you were bidding $2 on target “x” a couple of weeks ago before running your goal on Perpetua, the engine will not be able to get access to that data.
That’s why we give a lot of importance to the initial Target ACoS of your goal.

Suggested Actions

Review the historical bid of that set of targets, and apply a bid adjustment. The bid adjustment to consider could be a fixed bid adjustment for the next 30 days that would match your historical. The engine will learn from this “experiment” by considering the performance of the target during the 30-day experiment.

What happens to bids when a bid adjustment expires?

Our engine takes the data from these experiments and trains the algorithms for your specific account, using both good and bad results. All the experiments can be adjusted on the dedicated "Experiments tab", where you can change the bid, adjust the duration, or stop the experiment.

You can choose a fixed bid or a bid multiplier which will be optimized for a set duration that you choose. During that time, you can analyze the impact of the change and its results within the original goal and in the Experiments tab.

With the ad engine's continuous learning, it will take all the data collected during the experiment to further improve its machine learning and better optimize goals.

Article last updated June 2024. If you find this information to be out of date, please contact hello@perpetua.io.