Quick Links
In Amazon advertising, where your ad appears is often just as critical as which keywords you bid on. Amazon divides your ad traffic into three main placement zones: Top of Search (ToS), Rest of Search (RoS), and Product Pages (PP).
When one of these placements begins to bleed your budget with a skyrocketing ACoS, it drags down the performance of the entire campaign. Manually hunting down these discrepancies and calculating the perfect Bid Adjustments is tedious and mathematically exhausting.
We have automated this entire process. ThinkAd’s new Campaign Placements Recommendations module scans your campaigns daily, isolates underperforming zones, dials down their visibility, and redirects your hard-earned budget to the placements that actually drive profitable sales.
1. How the Algorithm Spots Hidden Problems
The system doesn't make hasty decisions based on minor market fluctuations. To trigger a recommendation, a campaign must first pass a strict data-volume qualification check:
Campaign Age: Active for at least 30 days from its start date.
Traffic Volume: Minimum of 100 cumulative clicks over the last 30 days.
Placement Share: The specific underperforming placement must account for at least 5% of the campaign's total traffic (to rule out statistical noise).
The "Underperformance" Logic
Every 24 hours, the algorithm analyzes 30 days of historical data and calculates a critical efficiency threshold for each placement. The formula is straightforward: Bad Threshold = Campaign Average ACoS × 1.5.
If a placement's specific ACoS exceeds this threshold, it is flagged as a candidate for optimization.
Example: Your campaign's average ACoS is 34%. The critical threshold is set to 34 × 1.5 = 51%. If your Product Pages placement is running at a 55% ACoS, the system flags it as a major anomaly.
Next, the algorithm singles out the worst-performing placement (the one with the highest overspend), calculates a safe target CPC for it, and proportionally recalculates the desired bids for the remaining healthy placements. This backend math is then translated into clear, actionable percentage changes (Placement Adjustments) in your dashboard.
2. Managing Recommendations in the UI
Whenever the system detects a placement anomaly, a dedicated lightbulb indicator will appear in your Campaigns table.
Hovering over or clicking the indicator opens a tooltip card containing a concise summary: which placement is underperforming and the exact recommended percentage adjustments (+0%, +35%, +84%, etc.) required to fix the balance.
You always maintain full control over your strategy via two simple actions:
Scenario A: You click "Apply"
The system instantly pushes the new placement coefficients directly to your Amazon Advertising console. The campaign is then placed into an active monitoring phase.
Scenario B: You click "Dismiss"
If you prefer to keep your current settings (e.g., you are intentionally running a specific test or possess external context the system doesn't know), clicking Dismiss closes the card.
The system will not spam you with the same recommendation the next day. It respects your decision and remembers the dismiss event.
The placement recommendation for this campaign goes to "sleep" for 30 days.
⚠️ The campaign will only re-emerge from this status earlier if you manually modify its placement adjustments or click the Re-evaluate button in the ThinkAd UI.
3. Protection Against the Yo-Yo Effect (Frozen Status)
A common concern among seasoned sellers is: “What if we suppress traffic on a bad placement, its ACoS drops and looks great after a month, the system thinks it's healthy again, reactivates it, and it just bleeds my budget all over again?”
To prevent this cyclical trap, our algorithm includes a built-in safety state called Frozen.
If, after applying our placement recommendation, the campaign shows stable or improved overall performance across a 30-day monitoring window, the system automatically assigns a FROZEN status to the campaign.
This means: “The strategy worked, and the ideal placement balance has been locked in. We won't touch it.” While frozen, the placement algorithm halts further adjustments to those percentages, allowing ThinkAd to focus 100% of its optimization power on fine-tuning your individual Keyword Bids. The campaign will remain frozen unless you manually change the placement adjustments yourself.
4. Budget Insurance: Automated Rollback
Amazon's advertising landscape is dynamic. Sometimes external variables—like aggressive competitor price-slashing or sudden seasonal shifts—can disrupt even the most mathematically sound strategy. To protect your capital, we built a fallback insurance mechanism.
Once a recommendation is applied, ThinkAd performs strict performance audits lasting 30 days.
If, during these checks, the system detects that your overall campaign ACoS has actually increased by 10% or more compared to its original baseline, the algorithm triggers an automatic alert:
In your dashboard, a Rollback card will immediately appear:
⚠️ Rollback recommended: ACoS is trending up
The latest placement updates did not yield the expected results. To prevent further overspend, we recommend reverting to your previous values.
Clicking Rollback instantly resets your placement percentages to their previous state. The campaign is then put into a 30-day Cooldown phase, giving it ample time to stabilize before the system even considers running a placement scan again.
5. Why Did My Keyword Bids Drop? (Bid Compensation)
After applying a placement recommendation, you might notice that the base bids on your individual keywords within that campaign have automatically decreased. This is not a bug—it is a vital safety mechanism.
When you increase a placement modifier (for example, setting Top of Search to +84%), Amazon multiplies your base keyword bids by that percentage for that placement. If ThinkAd left your base keyword bids untouched, your real, effective cost-per-click (eCPC) would skyrocket, violently draining your daily budget.
To prevent this budget shock, the moment placement adjustments are updated, the algorithm calculates a Weighted Bid Compensation Factor based on click volume. The system then gently lowers the base keyword bids proportionally (while ensuring they never drop below Amazon's absolute minimum of $0.02).
Article last updated July 2026. If you find this information to be out of date, please contact [email protected].