Setting up your campaign effectively is crucial to achieving your goals. By following these key best practices from the start, you can lay the foundation for long-term success.

Start with a broad number of SKUs in your campaign

Include as many relevant SKUs as possible in your initial campaign (at least 10 SKUs if possible). This gives the algorithm a wide dataset to learn from and helps it identify which products are most likely to perform well within the auction. Narrow campaigns limit learning and can slow down performance.

Let the campaign run for at least 30 days

Avoid making changes too early. The system needs time to gather data to optimize. A longer run time allows the algorithm to stabilize and make smarter bidding decisions.

Use auto pacing

Auto pacing adjusts your daily spend based on traffic and opportunity. This helps you avoid overspending on slow days and ensures you’re competitive when traffic spikes, like during sales events or holidays.

My Ads offers three optimization models. Select the option that best aligns with your business goals.

• Click Optimizer: Use this if your goal is to drive traffic to your product pages. It’s ideal for new product launches or increasing brand visibility.
• Conversion Optimizer: This model focuses on getting more orders at the lowest cost per sale. It’s great for sellers focused on efficiency.
• Revenue Optimizer: This model aims to maximize return on ad spend (ROAS). It’s best for mature campaigns where profitability is the top priority. Choosing the right model helps align the bidding strategy with your goals. You can switch models later as your strategy evolves.

Successfully implementing a new optimization model requires a thoughtful approach and a bit of patience. Here are key tips to help you manage this implementation effectively and get the most out of your campaigns.

• Wait before making changes: After launching or adjusting your campaign, wait at least three to four days before evaluating performance. This gives the system time to adapt and avoids reacting to short-term fluctuations.
• Review performance weekly: Check your campaign once a week to identify trends. Look for top-performing SKUs, underperformers, and opportunities to adjust bids or budgets.
• Increase bids on high-performing SKUs: If certain products are converting well, consider increasing their bids. This gives them more visibility and can lead to better results.

Understanding how optimization models interact with ad serving logic unlocks more efficient and impactful campaigns. Here are common examples of optimization models and how they power real-time decisions.

Click Optimizer

1. A shopper visits BestBuy.com to buy a camera.
2. They type “camera” in the search bar.
3. Two products with the same relevance score enter the auction.
4. Camera A is bidding at $0.80 and Camera B is bidding at $0.75.
5. Camera A wins the placement because it bid higher.

Conversion Optimizer

1. A shopper visits BestBuy.com to buy a camera.
2. They type “camera” in the search bar.
3. Two products with the same relevancy score enter the auction.
4. Camera A is bidding at $0.80 and Camera B is bidding at $0.75.
5. Based on historical shopper data, the algorithm knows Camera A has a 5% predicted conversion rate and Camera B has a 6% conversion rate.
6. Camera A: $0.80 x 2% CTR X 5% = .0008 and Camera B: $0.75 x 2% CTR x 6% = .0009.
7. Camera B has a higher effective CPC than Camera A, so it wins the placement.

Revenue Optimizer

1. A shopper visits BestBuy.com to buy a camera.
2. They type “camera” in the search bar.
3. Two products with the same relevancy score enter the auction.
4. Camera A is bidding at $0.80 and Camera B is bidding at $0.75.
5. Based on historical shopper data, the algorithm knows Camera A has a 5% predicted conversion rate and Camera B has a 6% conversion rate. Camera A costs $60 and Camera B costs $40.
6. Camera A: $0.80 x 2% CTR X 5% x $60 = .048 and Camera B: $0.75 x 2% CTR x 6% x $40 = .038.
7. Camera A has a higher effective CPC than Camera B, so it wins the placement.

CPC bidding plays a pivotal role in determining your campaign’s visibility, competitiveness and overall efficiency. These recommendations offer a framework for smarter bidding decisions.

Start bids at 20% above the floor

For campaigns using the Conversion or Revenue Optimizer, begin with CPC bids that are about 20% higher than the minimum (floor) bid. This helps your ads gain traction early and ensures they’re competitive enough to win impressions.

Adjust CPCs gradually

Make changes in small increments, typically 5%-10% at a time. This allows the algorithm to adjust smoothly and avoids disrupting performance.

Use multiple levels of control

You can set and adjust CPC bids at different levels:

• Ad Group level: Useful for managing groups of products or categories.
• SKU level: Ideal for fine-tuning bids on individual products.
• Keyword level: Helps you focus on high-performing search terms.

Increase CPCs to improve delivery

If your campaign isn’t spending its full budget, increasing CPCs at the ad group level can help improve delivery and visibility.

Optimize for ROAS

To hit your return on ad-spend goals, monitor performance by SKU, placement and keyword. Adjust bids where it is needed to shift spend toward the most efficient areas.

A strong keyword strategy is essential to reaching the right shoppers at the right time. These recommendations will help you strike the ideal balance between automation and customization.

Harness the automated keyword model

The model automatically assigns over 200 keywords to each product, achieving an average relevancy rate of 92%, which is substantially higher than the 39% typically seen with manual keyword selection. This automation ensures a broader, more accurate reach, especially valuable when historical keyword data is limited.

Let the optimizer maximize your investment

The optimization engine intelligently allocates your budget across the top-performing keyword and SKU combinations. This ensures your spend is continuously directed toward the most effective pairings; no manual adjustments are required.

Expand reach with positive keywords

To scale beyond the automated model, you can manually add positive keywords. This is ideal for targeting specific search terms that align with your objectives or promotional strategies.

Refine targeting with negative keywords

Improve campaign efficiency by adding negative keywords based on low performance, poor ROAS or strategic exclusions. This helps eliminate spend on irrelevant or underperforming queries.

Control CPCs at the keyword level

Manually adjust CPC bids at the keyword level to boost investment in high-performing terms or reduce exposure on underperforming keywords. This level of control is especially useful when fine-tuning campaigns to meet ROAS or budget goals.