Are you “Neglectful Neil,” ignoring your bids until it’s too late? Or maybe “Anxious Andy,” panic-adjusting every time ACOS creeps up by a percentage point?
If you are treating every keyword, match type, and campaign with the same broad strokes, you are likely leaving money on the table—or worse, sending your account into a visibility death spiral.
We explore the traditional “good ACOS = bid up, bad ACOS = bid down” philosophy. From “Hero” products to “Bleeders,” and from competitor conquesting to defensive branding, discover why advanced bidding requires a custom, layered approach that goes far deeper than the Amazon Ad Console default metrics.
Table of Contents
The “Death Spiral” & The Case for Classification
If you have been managing Amazon PPC for any length of time, you know the drill.
You log in, filter by high ACOS, and start cutting bids. It feels productive. You are “optimizing.”
But according to our guest Priscilla Wagner, this standard operating procedure is often a trap.
Priscilla describes a phenomenon known as the “Visibility Death Spiral.” When you relentlessly optimize solely for ACOS (Advertising Cost of Sales), you might achieve a profitable ad spend, but you often suffocate your total sales in the process. You bid down to save margin, you lose impression share, your organic rank slips, and suddenly, your “efficient” account is generating half the revenue it used to.
To escape this spiral, you need to stop viewing keywords as simply “Good” or “Bad.” You need to start viewing them through a Classification System.
The default Amazon Ad Console gives you raw metrics: Clicks, Spend, Sales. But it doesn’t give you context.
Priscilla’s approach involves exporting that data (via bulk files) and creating “calculated fields” to assign a specific identity to every target.
Here is the breakdown of the primary classifications used to categorize keyword behavior:
The “Heroes”
These are the heavy lifters. Following the Pareto Principle (80/20 rule), these are the targets generating the vast majority of your revenue.
Increase the bid.
If it’s a Hero, you want to defend that placement aggressively. You are not trying to save pennies here; you are trying to own the shelf.
High Volume, Unprofitable
This is the category that scares most sellers. These keywords have high spend and an ACOS well above your target (e.g., 100%+).
Do not kill them immediately.
These terms often drive the volume required to maintain your organic ranking. Instead of pausing them, analyze what percentage of your total sales they influence. They may be necessary loss leaders.
Bleeders & Extreme Bleeders
These are the targets that spend money but return zero value.
Implement strict thresholds.
If a keyword hits a calculated threshold—say, 20 clicks with 0 orders—it gets automatically flagged as a “Bleeder” and the bid is slashed. This prevents emotional decision-making.
The “Good CTR” Anomalies
Not all non-converting keywords are trash.
Good CTR, Low Clicks: This is an opportunity. The market wants this product, but your bid is likely too low to get sufficient impressions.
Good CTR, High Clicks (No Sales): This is a red flag. It usually indicates a listing issue (price, review count, or images) rather than a traffic issue.
A “Hero” for a brand doing $10M a year looks very different from a “Hero” for a brand doing $100k.
The magic isn’t in the spreadsheet itself; it’s in customizing these definitions to match your specific product lifecycle.
The Power of Intentionality & Context
The final layer of this advanced bidding protocol is perhaps the most critical because it requires you to step back from the raw numbers and ask a fundamental question about why a campaign exists.
This is the concept of intentionality.
However, a sale generated from a competitor’s product page is fundamentally different from a sale generated by someone typing in your brand name.
If you asked any brand owner how much they would be willing to pay to steal $10,000 worth of market share from their biggest rival, the answer would likely be much higher than what they would pay to convert a customer who was already looking for them.
This implies that your bidding strategy cannot be uniform across the account.
If you apply a flat 30% ACOS target to everything, you will inevitably overspend on branded terms that should be much cheaper, and you will choke off your competitor conquesting campaigns because they naturally run at a higher cost.
A “Brand Defense” campaign is held to a strict efficiency standard, while a “Competitor Conquesting” campaign is given permission to run at a higher ACOS because the goal is aggression and market share acquisition, not immediate profit.
This logic extends beyond just campaign types and into the operational reality of the business.
The most sophisticated version of this bidding formula incorporates external data signals, such as inventory levels. If a product is running low on stock, the algorithm doesn’t just wait for the inevitable stock-out; it applies a “soft pause” or a calculated bid discount to slow down velocity without completely killing the campaign’s history.
Conversely, if a product is in a launch phase, the “Hero” logic overrides efficiency targets to push for organic ranking.
The Macro View — Escaping the Spreadsheet Tunnel Vision
The danger of building a sophisticated bidding machine—one with hero tiers, match type cross-sections, and inventory logic—is that you can easily get lost in the weeds.
When you are staring at thousands of individual rows in a bulk file, it is impossible to see the bigger picture. You might be optimizing individual keywords correctly according to your formula, but is the account actually moving in the right direction?
To solve this, Priscilla introduces a crucial step in her workflow: The Directional Check.
Before uploading any bulk file back to Amazon, she doesn’t just look at the individual bid changes; she looks at the averages.
By creating a summary table (often a pivot table) of her proposed changes, she groups the data by her performance categories. This allows her to answer the “sanity check” questions that an algorithm might miss.
For example, she looks at the “Hero” category as a whole. If her formula is suggesting that the average bid for her “Hero” products should decrease by 5%, she knows something is wrong. Heroes are meant to be aggressive; unless there is a massive stock issue, that category should generally see flat or increasing bids. Conversely, if she looks at the “Bleeder” category and sees that the average bid change is positive, she knows the logic is flawed.
This summary view acts as a safety valve. It prevents the “death by a thousand cuts” scenario where a mathematical error in a formula quietly destroys campaign performance over a few weeks. It ensures that the strategy is actually being executed by the tactics.
This process also highlights that a bidding protocol is never truly finished.
The market changes, CPCs rise, and Amazon releases new ad types. A rigid system eventually breaks. The most successful advertisers are those who maintain a “Change Log,” treating their bidding strategy like software that needs constant updates, patches, and new feature releases to stay compatible with reality.
The Placement Puzzle & The Limits of Automation
Even with a complex Excel model driving the majority of bidding decisions, there is one area that remains stubbornly difficult to fully automate: Placements.
In the Amazon ad ecosystem, a bid is not just a single number; it is a base from which you can apply multipliers to appear at the “Top of Search” or on “Product Pages.” During the discussion, Priscilla admits that while her system is highly advanced, placement optimization often requires a manual touch, particularly for her “Hero” products.
The logic here is nuanced. If a Hero product is converting exceptionally well at the Top of Search, the algorithm might suggest lowering the base bid to save money, which would inadvertently kill the placement performance.
To counter this, Priscilla uses a set of “hypotheticals” within her logic. She looks for scenarios where Top of Search sales volume is high and efficiency is within range.
In these cases, the protocol might deliberately lower the base bid while simultaneously increasing the placement multiplier.
This “seesaw” approach allows her to target the specific real estate on the page that is driving revenue, rather than just bidding blindly on the keyword.
However, this is where the human element becomes irreplaceable. Unlike keyword bids, which can be adjusted by percentage points across thousands of rows based on math alone, placement modifiers can drastically swing spend.
This entire process—from classification to match type analysis to placement checks—relies heavily on the use of Bulk Files.
This isn’t a strategy you can execute by clicking inside the Seller Central dashboard. It requires downloading the raw data, processing it through these custom Excel “drivers,” and re-uploading it. It is a heavy data-lifting operation that, as joked about in the episode, requires a very fast computer. But for those willing to deal with the spreadsheets, it offers a level of competitive advantage that standard automation tools simply cannot match.
That feeling when Amazon PPC data is easy to read.
The Uncharted Territory & Continuous Evolution
The most intimidating aspect of building a custom bidding engine is the realization that it is never truly finished.
The next frontier for this logic involves layers that go beyond simple keywords. There is the rising importance of B2B targeting, which requires a completely different set of valuation metrics since business buyers behave differently than retail consumers.
Then there are the new “Audience” levers Amazon is pulling, moving targeting away from just “what users search” to “who users are.” And finally, there is the Amazon Marketing Cloud (AMC), which promises to unlock data that has previously been invisible.
Automation requires a critical mass of predictable data to work correctly. Until a specific feature—like B2B targeting—becomes a significant revenue driver for a specific client, it is better to manage it manually than to let a formula make guesses based on thin data.
This reinforces the core philosophy: use automation to scale your proven strategies, not to test new ones.
Conclusion
If there is one overarching theme to take away from this conversation, it is the concept of depth.
Most Amazon sellers are operating at the surface level, reacting to the metrics that Amazon presents on the default dashboard. They see an ACOS number, and they react.
But the “Advanced Bid Optimization” that Priscilla Wagner demonstrates is about seeing the matrix behind those numbers. It is about understanding that a 40% ACOS on a competitor term is a victory, while a 40% ACOS on a branded term is a failure. It is about knowing that a Broad Match keyword needs a longer leash than an Exact Match keyword.
It is about realizing that your inventory levels, your organic rank, and your product lifecycle should dictate your bid just as much as your conversion rate does.
You do not need to be an Excel wizard or a data scientist to adopt this mindset. You simply need to stop treating your account as a monolith. Start by classifying your products. Define your heroes. Separate your conquesting dollars from your defensive dollars. If you can bring that level of intentionality to your bidding, you are already miles ahead of the competition who are just blindly chasing a lower ACOS.
The PPC Den Podcast
SUBSCRIBE
- 🦡 GET AD BADGER ➡️
- Find Michael Erickson Facchin — Ad Badger’s CEO and soul, PPC whisperer, and your go-to guide through the wild world of Amazon advertising
- Our guest today is Priscilla Wagner of Relentless B2C
- The Simplified Q4 Bids & Budget System for Amazon Sellers
- The One Activity That Will Fix Your ACOS in 2025
- One Keyword, 8 Tools. Accuracy for Search Volume & Sales on Amazon
- How to Optimize Minimum and Maximum Bids in Amazon PPC to Improve ACoS
- 📚 Unlock our FREE comprehensive Amazon Marketing Playbook
- Subscribe to our newsletter
- Transform your workflow with our comprehensive Amazon Advertising checklist
- Review all our show notes
Special thanks to Sofiia Podash, Pedro Moreno, Priscilla Wagner and Michael Erickson Facchin for the production of this blog.
Check out more from our blog:
Mastering Amazon PPC With a Custom Excel-Powered Bidding System
