Skip to main content
All CollectionsPredictive AI Bidding GuidePredictive AI Bidder FAQ
Decoding Predictive AI Bidding: A Guide to Optimizing Your Advertising Strategy
Decoding Predictive AI Bidding: A Guide to Optimizing Your Advertising Strategy

The Bidder calculates bids for biddable targets and ad items on both Amazon and Walmart to optimizing your advertising efforts.

Updated over 6 months ago

When bidding on a target, the seller with the highest bid and best ad will win. After winning, the seller’s ad is placed in search results or a product detail page when a customer’s search maps to the target. Then, if the customer clicks on the ad, the seller pays for that click, which will hopefully lead to a sale.

How does the Bidder calculate the optimal bid for your business?

To understand how the Bidder works, it's essential to grasp the fundamentals of the second-price auction. Because your ads will only show up if you win a spot in the auction, figuring out the optimal bid for every keyword and target is very important so that you don't underspend or overspend on advertising. You can read more about how second-price auctions work, but here are the key takeaways:

  • Every time a shopper searches on Amazon or Walmart, Amazon and Walmart both run a second-price auction to determine which ads are shown and in what order.

  • In a second-price auction, all participants submit their bids simultaneously, without knowing the bids of the other auction participants.

  • The advertiser who has submitted the highest bid wins the auction, but only pays the next-highest bid as a price.

  • On Amazon and Walmart, the second-price auction considers the bid AND relevance of the search term to your ad. This means that relevance can sometimes outweigh a higher bid, leading to a more cost effective placement where budgets go further and profit is maximized.

  • The optimal bidding strategy in a second-price auction is to bid the true value of the ad placement to the advertiser.

  • This strategy ensures that advertisers pay an amount equal to what the placement is genuinely worth to them.

The journey of the bidder remains the same for targets on both Amazon and Walmart: starting with a starter bid, entering the discovery stage, and then computing the value bid.

Let’s walk through each stage of the bidder’s journey.

What is the Starter Bid?

The first bid calculated for a target is known as the ‘starter bid’, when there is not enough click history to estimate the value of a click.

The starter bid estimates click value using Teikametrics proprietary AI based on factors such as campaign average conversion rate and product average payoff per order. This will also be used for targets that revert back to a 'low data' state when not having click traffic for multiple weeks. In both instances, the starter bid kicks off the discovery stage.

What is the Discovery Stage?

When starting out with bid automation, launching new campaigns, or adding new targets, the bidder is in the discovery stage after the starter bid is set. During this stage, the bidder will slowly raise the bid for the target until it has enough clicks to determine the true value of a click. The objective here is for the bidder to buy some clicks so it can learn from real data what the true value of a click is for the specific target. In auto campaigns, bids are calculated at the ad group level and in manual campaigns at the target level.

On average, the discovery stage can take 15 days. The timeline depends on the relevancy of the targets, the amount of sales data, and the set ACoS Target. A higher ACoS Target will speed up the initial discovery process, but may also incur additional ad spend and a higher ACOS. Once we have determined the true value of the bid after having gathered enough data, we transition to the value-based bid flow.

What is the Value-Based Bid?

When bids have entered the value bid flow, we begin calculating bids based on changes in performance. The equation used is:

(payoff per order) x (conversion rate) x (ACoS Target) = Value Based Bid

Any changes in 1) conversion rate, 2) payoff (attributed ad revenue) per order or 3) ACoS Target (the seller’s willingness to spend will lead to changes to bid value.

For the first-price auction on Walmart, bid shading is then applied to lower the bid to a more optimal value.

What factors govern bid changes in the Value-Based Bid?

Changes in Conversion Rate:

  • Increased: If the conversion rate (from clicks to orders) goes up, then the bidder raises the bid to take advantage of it.

  • Decreased: If the conversion rate goes down, then the bidder lowers the bid to account for the decreased value per click.

Changes in Payoff per Order:

  • Increased: If payoff per order increases (a price increase or the average number of units per order increased), then bids increase.

  • Decreased: If payoff per order decreases (reduced price, lightning deal), then bids decrease because the potential payoff (ROI) is lower. For example, if the price was originally $15, an ACoS Target set at 10% indicates a willingness to spend up to $1.50 on a click (assuming one unit per order). Reducing the price to $10, then indicates a willingness to only spend $1.00 per click.

Changes in ACoS Targets:

  • Increase in AcoS Targets = higher bids: Bidder performs more aggressively

  • Decrease in ACoS Targets = lower bids: Bidder performs more conservatively

Did this answer your question?