The Case for Bid Modifiers

When Amazon Ads rolled out bid modifiers in Q4 2024, the idea was simple but powerful: instead of setting one static bid per line item, you could now dial bids up or down based on six dimensions—audience, device type, operating system, geo-location, creative, and domain.

However, many savvy advertisers often avoid using it because of the operational lift and cumbersome means of execution. That’s where Gigi comes in.

Using the bid modifier API, Gigi analyzes thousands of rows of performance data into natural language decision summaries that surface exactly where your dollars will work hardest, plus one-click “accept” or “decline” recommendations to apply the modifiers. No manual guesswork, no drowning in line items and spreadsheets.

Lemon Perfect’s Summer Pressure Cooker

For Lemon Perfect, an organic, zero-sugar beverage brand, summer is their busiest season. July is a critical month, where peak demand meets intense competition, and every opportunity to win counts. Coming out of early-season testing in 2024, their full-service Amazon agency, Cartograph, wanted more granular levers to pull for their top-of-funnel display campaigns during this critical sales window. Enter: bid modifiers.

But testing bid modifiers wasn’t as simple as flipping a switch. To get a true read on performance, Cartograph partnered with Gigi to test the effectiveness of bid modifiers. Executing true A/B tests within Amazon Ads often a challenge. There are always variables like seasonality or inventory issues that can make comparing time periods uninsightful and certainly not a true A/B test.

We discovered the only way to do a true A/B test would be using AMC audiences. The team did this by assigning users into two perfectly balanced groups based on their user IDs:

• odd-numbered IDs formed the experimental group (receiving tailored bid modifiers)

• even-numbered IDs formed the control group (no bid modifiers applied)

From there, Cartograph duplicated every single line item in the campaign. One set targeted the control group, the other targeted the experimental group. The only difference? The experimental line items had bid modifiers applied, while the control group’s bids stayed untouched. Everything else: inventory access, creative, etc, remained identical.

The result was a clean, high-integrity test environment that allowed Cartograph to isolate the exact impact of bid modifiers, providing a clear view into how these adjustments could influence performance during Lemon Perfect’s most competitive season.

Finding the Sweet Spots

For the experimental group, Gigi analyzed June performance data and highlighted clear efficiency opportunities to Cartograph within its decision summary:

“Desktop users-particularly those on Windows and macOS- consistently delivered stronger engagement and conversion rates, justifying increased investment in these segments. Mobile performance varied, with Android phones generally outperforming iOS due to broader reach and cost efficiency, while iOS results were mixed depending on campaign context. Regionally, Florida and New York frequently stood out for high conversion efficiency, whereas California and Texas often faced higher competition and lower incremental returns. Bid modifiers have been strategically applied to prioritize top-performing device/OS and regional combinations, ensuring spend is focused where it drives the greatest impact on campaign objectives.”

From these insights, she then set forth the following recommendations for Cartograph to action:

The Results

After applying the bid modifications, the numbers told a clear story: while the experimental group only saw a 1.6x lift in spend and 1.25x lift in impressions, they dramatically outperformed across downstream metrics:

• 36x lift in assisted new-to-brand purchase users

• 44.78x lift in purchases

• 24.89x lift in total sales

• 15x lift for assisted ROAS

• 6x lift in detail page view rate

• 35.89x lift in purchase rate

And perhaps most striking: the experimental group achieved a CPA of $23 vs. $638 for the control group.

Note: We recognize these results are insanely strong, but we have tripled checked the output, and it’s real! While we don’t assume every customer will have these same results, we do believe this proves the benefits of applying bid modifiers. And for Cartograph, this experiment helped demonstrate to Lemon Perfect that by targeting the right device, OS, and geo combinations, the brand could boost conversions without chasing an impression spike.

The takeaway is simple: Don’t sleep on bid modifiers. With the right analysis and a little strategic tweaking, they can be one of the most impactful levers in your Amazon DSP toolkit.