Amazon PPC Incrementality: Which Ads Actually Drive New Revenue

The Organic Cannibalization Problem: When PPC Steals From Itself

On a mature account, the riskiest campaigns fall into three categories. Ordered by cannibalization risk, the following are the three campaign types most likely to cannibalize organic:

• Branded keyword Sponsored Products carry most of the risk. When you rank first organically and bid on your brand name, you’re cutting off buyers that were already coming to you. In most mature accounts, when you pause these campaigns, you get a partial or no drop in revenue and not the big loss agencies typically predict.

• Exact match Sponsored Products carry moderate risk. If your ASIN is already ranking in the top three positions organically for a keyword and you’re bidding on that same keyword in exact match, the only extra benefit you’ll get from the paid ad is the small increase in clicks when your organic listing and paid ad are showing at the same time. The increase is minimal once you’ve already claimed the top organic position, as most buyers were already clicking on you.

• Sponsored Display retargeting on existing customers is a situational risk. Retargeting buyers who purchased within the last 30 days has near-zero incrementality because they’re already in the funnel. Though, retargeting lapsed buyers in the 90-to-180-day window at the right margin threshold can be truly incremental.

How to Identify Cannibalization in Your Account

There are four signals worth a look before you run a formal test:

• Total advertising cost of sales (TACoS) is flat or up and ACoS is healthy. It's the clearest signal at the account level that ad spend is recapturing organic demand, not creating new revenue.
• You’re the dominant player in both organic and paid results for the same query. In this case, the paid part has very low incrementality on that term.
• There’s no meaningful drop in total ASIN revenue after pausing a campaign for 7 days.
• Spend is primarily hitting repeat buyers, indicated by a new-to-brand (NTB) order rate below 30% found in the Campaign Manager’s Brand Metrics tab. This number is based on Olifant Digital’s managed account data across categories in non-branded campaigns. It’s also category-dependent. For example, beauty and supplements will generally have a higher NTB rate than accessories or home goods. As such, use it as a guideline, instead of a fixed rule. That said, in most categories a non-branded campaign running below 30% is worth taking a look at as it’s mostly reaching repeat buyers.

The Cannibalization Calculation: What It Costs in Real Numbers

Cannibalization cost is basically the percentage of your ad spend that’s taking back sales you would’ve made organically.

Imagine a Sponsored Products campaign with branding that has been running for a month. You hit pause for a week, and revenue drops almost nothing. In some instances, organic traffic might increase as buyers are clicking your organic listing rather than your paid ad for the same search result. This means the campaign was spending money to win over the same buyers who would’ve found you anyway.

Your dashboard might say 4x ROAS, but the incremental ROAS is basically zero because almost all that revenue wasn’t incremental. The campaign was paying to regain buyers that were already coming in organically. The reported ROAS was measuring recapture efficiency, not new revenue generation.

This is why calculating your advertising cost of sales (ACoS) alone isn't enough. An account can look great in terms of ACoS, but run campaigns that cost more than they generate in real new revenue. A pause test will tell.

Why ACoS Hides Cannibalization and TACoS Surfaces It

Advertising cost of sales (ACoS) measures the ad spend divided by ad-attributed revenue. It has no visibility into organic revenue at all.

Aggressive bidding on branded terms may lead to an increased ACoS, but this can be offset by increasing organic buyers captured via a paid link while organic revenue stays flat or decreases. On paper, the account can look more efficient, but, in practice, the business increasingly depends on ads.

By contrast, total advertising cost of sales (TACoS) is calculated as ad spend divided by the sum of organic and paid revenue.  

At Olifant Digital, ACoS serves as a campaign-level efficiency metric, while TACoS functions as the account health metric. TACoS is where incrementality lives, not ACoS.

Reading the TACoS Trend: What Healthy Incrementality Looks Like

There are three TACoS scenarios that tell you where your account is at:

If TACoS is falling and total revenue is rising, ads are generating new revenue and stacking into organic rankings. This is the signal that Olifant Digital looks for in accounts three to 12 months after optimization.

If both TACoS and total revenue are flat, ads are holding their position but not producing new revenue. The investigation should start with organic rank stagnation and listing conversion rate.

If TACoS is up and total revenue is flat or down, ads are taking the place of organic decay. The worst case is when increased spend to compensate for declining organic revenue doesn’t fix anything. The root cause is often organic rank loss, e.g. CVR drops or competitor gains.

To illustrate this, you can refer to the work we completed for Onsen Secret. Its ACoS seemed under control, but TACoS was rising. This meant the ads were making up for the lack of organic performance, not creating new revenue.

We shifted reporting to TACoS per ASIN, honing in on the spend on proven high-converting products and optimized listings before scaling. After introducing TACoS-first reporting per ASIN, the skincare brand added $95,934 in monthly revenue.

💡Pro Tip: Go to Seller Central Business Reports and look for the Detail Page Sales and Traffic section. Check out organic sessions for every ASIN per week. If your TACoS is going up and your organic sessions are declining, your ads are compensating for a drop in organic traffic and not driving new sales.

TACoS by ASIN: Where the Real Signal Is

Account-level TACoS is a good health indicator, but it’s the ASIN-level TACoS where the real incrementality insight lives. For example, one account can have a healthy blended TACoS of 8%, while two ASINs are running at 35% TACoS and two others only at 4%. The blended number completely hides both situations.

To avoid this, Olifant Digital provides weekly TACoS per ASIN and alerts you to any ASIN with a TACoS that’s rising faster than total revenue growth over a rolling four-week period. This metric is the early warning sign for cannibalization or organic decay.

We do this with Olifant AI, our proprietary Amazon PPC and account management platform, which provides daily reporting so that senior specialists can act before cannibalization worsens.

For example, for Spade to Fork we implemented product-level PPC with TACoS tracking at the ASIN level. This meant we could scale based on the ASINs that were improving TACoS. This way, the organic gardening brand increased its revenue by 46% in 44 days, while enjoying an ACoS reduction of 19%. Ad sales also grew 132%, driven by high-incrementality campaigns.

The Incrementality Formula: How to Calculate What Ads Actually Contribute

Incremental revenue is revenue that wouldn’t have happened without the ad. In other words, the amount of sales attributed to the ad that wouldn’t have been realized through organic discovery.

To calculate it, look at your revenue during a normal period vs revenue when you stop a particular campaign. This difference shows you how much that campaign actually contributed to new sales.

To do this, you need a good organic baseline and create a clean campaign structure so you don’t get mixed signals from multiple campaigns at the same time. This is why Olifant Digital’s 1-1-1-1 method is so important. Messy account structures make it impossible to isolate what’s really driving the revenue change.

Incremental ROAS vs Reported ROAS: The Gap Most Accounts Won't Look At

Incrementality is very different between different types of campaigns. For example:

• Exact match non-branded category keywords are the best-performing keywords because you’re reaching buyers who have never heard of your brand. Nearly all of that revenue is truly net new.
• Sponsored Brands Video is also doing well on non-branded searches because those buyers are discovering you for the first time.
• Sponsored Display is a powerful way to target competitors because you’re intercepting shoppers who are seeking alternatives.
• Sponsored Products on branded accounts that have a strong organic rank do the worst because you’re mostly retaking buyers that would’ve found you anyway via organic search.

Which ASINs Are Worth Testing (Prioritization Framework)

Not all ASINs require an incrementality test. Prioritize them based on the following criteria:

• High ad spend and strong organic rank (start testing these as they have the highest risk of cannibalization)

• Branded campaigns on mature ASINs (12+ months’ history, established review base, and increasing branded search volume)
• TACoS flat or increasing for 8+ consecutive weeks with consistent ad spend

💡Pro Tip: Don’t test new ASINs that are less than 6 months old, have fewer than 50 reviews, and a weak organic rank. PPC is giving it ranking velocity, which will build into organic growth and the incremental value is really high at this stage.

How to Run an Amazon PPC Incrementality Test (Step by Step)

Step 1: Select the Test ASIN and Establish a Baseline

Select only one ASIN (don’t test with multiple ASINs at the same time) using the prioritization criteria described in the previous section.

Pull the past four weeks’ daily data for:

• Total ASIN revenue (organic and paid)
• Organic sessions from Seller Central Business Reports instead of Campaign Manager Ad spend on the campaign being tested
• ACoS

Don’t initiate the pause during a promotional event, holiday week, or within seven days of a price change, as baseline data needs to be established from a steady-state period.

Step 2: Define the Test and Control Conditions

Pause the specific campaign. It has to be a single campaign focused on the target ASIN.

This is why the 1-1-1-1 method is so important. In a well-structured account, you can isolate one branded exact match campaign on one ASIN and have zero impact on any other traffic.

If you pause, only this campaign will stop running. Ad spend naturally falls to zero. Testing one campaign’s incrementality, rather than the total PPC impact, allows all other campaigns for this particular ASIN to continue running.

Run the test for a minimum of two weeks. Seven days aren’t enough, as the 48-hour attribution lag means that two of those days are noise. That sais, 21 days are ideal for accounts with daily revenue variation.

Step 3: Initiate the Pause

Stop the campaign at the beginning of a Monday to normalize against weekly traffic patterns in the baseline comparison.

Don't make any decisions based on the first seven days of data due to attribution lag. Amazon’s 48-hour attribution delay means early data is unreliable. Checking it on a daily basis will only create false signals and will lead to premature test termination.

On day seven, get the total ASIN revenue from Business Reports, along with organic sessions and conversion rate for reference, but don’t take any action yet. Only take action after two weeks.

Step 4: Measure the Revenue Delta

You should compare the test period to the four-week baseline after 14 to 21 days and determine whether any seasonal index adjustment is needed.

To work out your revenue delta, take your daily average revenue from the baseline and minus your daily average revenue from the test period.

For example, if your baseline was $1,200 a day and your test period was $1,160 a day, your revenue delta is $40 a day, or approximately $280 over a seven-day window.

Also, you need to monitor organic session count while the campaign is paused. The most underutilized data point in cannibalization testing is organic session count during the pause.

If organic sessions go up while the campaign is off, your paid ads are suppressing organic click-through rate (CTR) by taking up the same search real estate.

Step 5: Interpret the Result and Make a Decision

There are three scenarios that can happen. Here’s what Olifant Digital recommends for each:

• Revenue drops significantly (more than 15% of the ad-attributed revenue): Reinstate it and think about whether you can get more efficient with match type, bid levels or negative keywords.

• Revenue down slightly: The campaign delivers partial incrementality, and it’s worth keeping it at a lower budget. Reduce the spend by 30 to 50% and then re-test in 60 days to see if the lower budget still delivers incremental value.

• Revenue flat or bounced back within a week: The campaign is mostly eating into organic sales, meaning there’s no incrementality. You can either kill it altogether or reduce it to a bare-bones defense budget of $200 a month to keep up with the search results. Reallocate the freed budget to non-branded exact match or Sponsored Brands Video campaigns with higher incremental ROAS.

💡Pro Tip: If you already have the top three organic positions for your branded terms, turning off your branded ads probably won’t hurt your organic rank. However, if your organic rank is weaker for category keywords, your ads might help to hold that position, making pausing riskier.

Does Amazon's Multi-Touch Attribution Beta Solve the Incrementality Problem?

Amazon’s Multi-Touch Attribution (MTA) in beta is now live in the US and many sellers are asking if it means the end of manual incrementality testing.

In short, it doesn’t. That said, it does give you a better picture of how your ads are working as a whole.

MTA assigns credit to all ad touchpoints in the 14-day window, rather than giving full credit to the last ad a buyer clicked on. This way, you can see how upper-funnel formats like Sponsored Brands and Sponsored Display are contributing to conversions.

What MTA can't tell you is whether those conversions would’ve happened without the ads. The cannibalization problem is invisible in MTA data since it considers only paid touchpoints and no benchmark is made with an organic baseline.

Olifant Digital uses MTA to help optimize budget decisions on Sponsored Brands and Sponsored Brands Video, but it doesn’t replace the pause test. For ASINs that are already ranking well organically, the pause test is still the only way to measure true incremental ROAS.

How Olifant Structures PPC for Incremental Growth

Olifant Digital classifies every campaign by incrementality tier before making any budget recommendation.

Tier 1: High Incrementality Spend

For category keywords where your organic rank is fourth or lower, exact-match paid reaches buyers who wouldn’t otherwise find you. Every sale is a new sale.

Sponsored Brands Video on unbranded queries converts buyers who are seeing your brand for the first time and, by definition, drives incremental conversions.

Use Sponsored Display and Sponsored Products for competitor product page targeting. These buyers are actively considering a competitor, meaning you can pull them into your funnel rather than recapturing existing demand.

Tier 2: Moderate Incrementality

Non-branded phrase matches in discovery campaigns capture buyers in your category. Some traffic is new while some overlaps with organic.

Auto campaigns help you discover new keywords by generating new search terms and reaching buyers who aren’t yet being targeted. That said, the broad nature of auto targeting means that not every click is a truly new buyer.

Sponsored Brands headlines are category queries that help you get brand awareness for shoppers in your category. Since conversions mix with organic results, the strength of your organic presence determines incrementality.

Tier 3: Low Incrementality

This tier is for branded exact matches on mature ASINs with strong organic rank, since most people searching your brand name would’ve found you anyway organically.

At Olifant Digital, we run these campaigns on a bare-minimum defensive budget unless a competitor is actively bidding on your branded terms. In this case, the spend is incremental as it’s protecting your position, not recovering organic buyers.

Tier 4: Near-Zero Incrementality

This tier addresses two scenarios in which ad spend generates virtually no incremental revenue.

The first is branded exact match, where your organic rank is already in position one and there are no active competitors bidding. There’s no meaningful traffic to protect and almost every buyer clicking the ad would’ve converted through the organic listing anyway.

The second is retargeting existing customers within 30 days. These buyers already know your brand and have recently purchased so the ad isn’t generating new demand. On accounts focused on profitable growth, the budget should be close to zero.

TACoS-First Reporting as the Incrementality Accountability Layer

Every week, Olifant Digital reports TACoS per ASIN, and this accountability layer demonstrates incrementality without requiring a monthly pause test. Incrementality isn’t a metric you check quarterly; it’s the framework that informs every budget allocation decision from day one.

This strategy is made operational at scale with Olifant AI, which monitors per-ASIN TACoS, organic sessions, and keyword coverage on a daily basis across 50+ accounts simultaneously. It surfaces incrementality signals the moment they appear so our senior specialists can act on them. No work is delegated below senior level so each of those signals is looked at by a specialist with at least seven years of experience.

If you’re spending $10K+/month on Amazon PPC and haven’t done an organic cannibalization test, you’re probably wasting some of that money recapturing revenue you would’ve made anyway.

Get a free marketing plan from Olifant Digital. We’ll identify your highest-risk campaigns for cannibalization and tell you exactly how your incremental ROAS appears across your account. Plus, all work after the free assessment is backed by a 60-day money-back guarantee.

Frequently Asked Questions

What Is Incrementality in Amazon PPC?

Incrementality measures whether your ads are truly creating new revenue or merely claiming sales that would’ve occurred anyway. It lets you shift the conversation from reported ROAS to actual business growth.

How Do I Know If My Amazon Ads Are Cannibalizing Organic Sales?

The easiest way is to compare your TACoS trend over time to your total revenue growth. If TACoS is rising but your revenue is stagnant, your ads are probably covering the gap left by declining organic performance, rather than driving new sales. To be sure that the revenue is truly incremental, conduct a short pause test on your highest-spending campaigns.

What Is a Good Incremental ROAS on Amazon?

A good incremental ROAS is one that pays for your fully loaded unit economics at the margin you want to scale. For non-branded category campaigns the threshold is higher than most branded benchmarks suggest because the customer is a true new customer. Below 1x means the campaign costs more than it’s generating in truly new revenue.

Should I Pause Amazon-branded Keyword Campaigns?

If a competitor is actively bidding on your brand name, branded spend is clearly justified. If not, a pause test will quickly reveal how much of that spend is incremental vs organic recapture.

How Does TACoS Measure PPC Incrementality?

TACoS gives insight into ad efficiency that ACoS can’t, because it looks at total revenue, not just ad-attributed revenue. If your total revenue is increasing and your TACoS is decreasing, that’s the best sign your ads are driving organic growth.

‍

‍