Anyone who steers their Amazon budget today purely by last-touch ROAS is driving a state-of-the-art car stuck in first gear. ROAS is a convenient metric for quick wins, but it is blind to causality. It doesn't tell you whether a sale would have happened without the ad anyway, how many touchpoints a new customer really needs, or when your budget is simply running into market saturation. In a competitive market environment, this "high-level guesswork" is a dangerous brake on growth.
Amazon Marketing Cloud (AMC) breaks through this limitation. As a privacy-compliant "clean room" environment, it gives us access to event-level data that goes far beyond what Campaign Manager can deliver. At Xingu Advertising GmbH, we use this infrastructure to map the invisible paths of the customer journey and turn your media planning from a reactive into a proactive discipline.
Customer loyalty & acquisition logic: analyzing the CLV and NTB gateway
A central KPI for sustainable brand growth is Customer Lifetime Value (CLV). While conventional reports only measure momentary revenue, AMC allows us to identify buyer segments based on their long-term profitability.
- Segmentation into loyalty buckets: we categorize buyers into segments, from opportunistic one-time buyers to loyal brand ambassadors. This analysis often reveals an uncomfortable truth: campaigns with a seemingly "poor" ROAS are often the strongest drivers of long-term brand value, because they acquire exactly the customers who repurchase organically for years.
- High-velocity entry points (NTB gateways): thanks to AMC insights at the ASIN level, we identify your "gateway products." When we can statistically prove that buyers of Product A have a significantly higher probability of switching to your premium segment within six months, we radically prioritize the new-to-brand (NTB) budget toward this entry point, even if the initial ROAS there is lower.
Timing & purchase cycles: optimizing Inter-Purchase Time (IPT)
Consumer behavior follows patterns that we make mathematically tangible by evaluating Inter-Purchase Time (IPT). A major problem in digital marketing is bad timing: advertise too early and you burn budget; come too late and you lose the customer to the competition. By analyzing the individual repurchase cycles across your entire assortment, we determine the ideal window for re-engagement.
Example: if the data shows that a product is used up on average after 45 days with daily use, we start serving ads precisely on day 38. This predictive timing turns advertising into a useful reminder at the ideal moment (point of need). We also use this data to build churn analyses: customers who exceed their usual IPT window are specifically targeted with re-engagement campaigns to prevent them from switching to the competition. Users aren't flooded repeatedly with the same messages, but instead experience a consistent customer journey across all touchpoints.
Channel synergies: media overlap and incrementality measurement
We answer the fundamental question of incrementality: how much additional uplift does Search generate when the customer already had awareness touchpoints via DSP beforehand?
- Halo effect analysis: we see exactly how many touchpoints are needed to convert a new customer.
- Keyword-to-ASIN mapping: we understand which generic search terms were triggered by display ads and later led to purchases of (non-advertised) halo products.
A critical deficiency in media planning is the cannibalization of Sponsored Ads by Amazon DSP. Both channels often claim the same sale for themselves, leading to a distorted measurement of success. AMC lets us resolve this media overlap. This understanding allows us to orchestrate budgets across channels so that Search and DSP reinforce each other instead of competing for attribution.
Budget efficiency: saturation curves and marginal return analysis
Marketing investments are subject to the law of diminishing returns. Beyond a certain point, increasing the budget no longer leads to more revenue, only to more expensive clicks and shrinking margins.
- Optimal frequency: we calculate the saturation threshold for each audience. At what contact frequency does attention tip over into "ad fatigue"? We find the "sweet spot" where your ad pressure achieves maximum impact.
- Data-driven frequency capping: using AMC insights, we implement capping that goes far beyond standard settings. We make sure your ad pressure is only applied where it can still change behavior. This protects your margin and prevents you from sinking budget into audiences that would have converted anyway (organic buyers) or are already oversaturated.
From analysis to activation: precise targeting via AMC Audiences
The most valuable analysis is useless if it doesn't translate into action. The decisive advantage of our AMC strategy is direct activation via AMC Audiences in Amazon DSP. We turn SQL results into surgically precise audiences that simply don't exist in the standard interface.
Examples of these "power audiences":
Cross-sell precision
customers who bought Product X but, despite a high IPT, have shown no interest in Product Y yet.
Churn prevention
buyers whose individual IPT window has already been exceeded by 10% (impending churn).
High-intent non-buyers
users who came via search keywords with extremely high relevance, viewed the product, but haven't converted yet due to DSP frequency saturation.
Case study insight: validation using the example of L'Oréal
Theoretical excellence only proves itself on a real portfolio. In working with brands like L'Oréal, we were able to show that this data-driven depth delivers a decisive market advantage:
Gateway optimization
by focusing on the right entry ASINs, long-term CLV rose measurably, because we acquired the "right" customers, not just the "cheapest" ones.
Sequential targeting
precise IPT timing reduced wasted reach in retargeting by a double-digit percentage.
Holistic attribution
demonstrating the media overlap led to a strategic reallocation of budgets, away from isolated campaigns and toward a synergistic full-funnel approach.
Behind the scenes: the DNA of an AMC deep dive
AMC analyses at this level require infrastructure and expertise that we have perfected over years:
Analytical engineering
every insight must be extracted from millions of event data points through custom-built SQL queries. We don't build static dashboards, we build dynamic query models.
Statistical relevance
we deploy deep dives specifically where the data volume (typically from a six-figure annual media budget upward) allows for robust patterns.
Historical lookback
we use lookback windows of up to 12.5 months to capture even complex seasonal cycles and the true long-term development of new customers.
Conclusion: AMC as a strategic compass in the Amazon ecosystem
In a market environment shaped by rising CPCs and increasingly complex customer journeys, AMC is the only way to keep the strategic upper hand. Anyone who understands today how synergies arise and when saturation limits are reached no longer steers their brand on hope, but with mathematical certainty.
At Xingu, we translate this complexity into your economic success. Are you ready to open the black box of your data?
FAQ – frequently asked questions about AMC analyses
Campaign Manager uses last-touch attribution based on aggregated data. AMC provides access to event-level data. That means we can link every individual touchpoint of a customer across different formats (video, display, search) and analyze the entire path to conversion.
Without knowing the saturation threshold, brands risk "overspending." AMC shows us exactly the point at which increasing the budget only drives up cost per acquisition (CPA) without significantly increasing volume.
No. The computing power and data storage take place within Amazon's AWS environment. As an agency, we handle the creation of the SQL queries and the interpretation of the results.
Yes. Via Amazon Attribution we can feed signals from external sources (e.g. social media or Google) into the data clean room. This makes it possible to measure how external impulses interact with Amazon-internal advertising measures and influence the final sale.