Context: where enterprise Amazon Ads programmes stop scaling
A single-marketplace seller at Rs 10-30 crore Amazon ARR runs Sponsored Products + Sponsored Brands inside Amazon.in, manages bids by hand or via the in-platform Bid Optimisation, watches ACOS weekly, and that is the entire programme. The team holds 50-150 active campaigns mentally; the brand-store + storefront are an afterthought.
Enterprise Amazon Ads programmes are structurally different. The enterprise operator runs Amazon ads across 3-7 marketplaces (Amazon.in + Amazon.ae + Amazon.sa + Amazon.com + Amazon.co.uk are the typical India-house international expansion shape), operates Sponsored Products + Sponsored Brands + Sponsored Display + Amazon DSP across upper-funnel + mid-funnel + lower-funnel + retargeting + new-to-brand growth, instruments measurement through Amazon Marketing Cloud (AMC) for closed-loop cross-channel attribution, holds a TACoS (Total Advertising Cost of Sale) discipline that supersedes the campaign-level ACOS number, and runs against operational ceilings that single-marketplace operators do not face.
This piece is the reference architecture and governance discipline Dcrayons applies on enterprise Amazon Ads engagements in 2026. It covers four areas the public Amazon Ads documentation under-foregrounds: the multi-marketplace portfolio strategy + bid-coordination pattern, the DSP + Sponsored Ads integration architecture, the Amazon Marketing Cloud measurement discipline, and the operational pattern past Rs 50 crore Amazon ARR.
Multi-marketplace portfolio strategy
Indian brands expanding past Amazon.in typically follow the path: India first, GCC (Amazon.ae + Amazon.sa) second for low-friction Arabic-language + ME expansion, UK third for English-language + post-Brexit GBP economics, US fourth for the demanding-but-highest-ceiling marketplace. Each marketplace operates as a separate Amazon Advertising Account; each requires its own seller account or vendor account in the corresponding Amazon Seller Central / Vendor Central.
The Dcrayons multi-marketplace portfolio pattern:
Brand-segmented account structure. Each marketplace gets its own Amazon Advertising Account; within each account, campaigns are organised by brand-line + by funnel stage. India-Beauty-Sponsored-Products is one cluster; India-Beauty-Sponsored-Brands is another; India-Beauty-DSP is another; same brand-line in the UAE account is a separate parallel cluster. Cross-brand campaigns are an anti-pattern; the algorithm cannot optimise mixed brands.
Catalog discipline upstream. ASINs in each marketplace are different ASINs (an Amazon.in ASIN and an Amazon.ae ASIN for the same product are not the same ID). The PIM (Akeneo, Salsify, or custom) holds the product master; ASIN mappings per marketplace are part of the catalog ETL. Without this discipline, you spend on out-of-stock ASINs or on ASINs that are not yet listed in the target marketplace.
Budget allocation logic at the portfolio level. Quarterly budget allocation across marketplaces uses these inputs: marketplace-level marginal-ROAS, marketplace contribution margin (different in each marketplace due to FBA + return-rate + currency + tax), strategic-growth signal (a new marketplace deserves over-investment in months 1-12 even at sub-economic ROAS to win share). The MMM or warehouse-MTA holds the marginal-ROAS by marketplace; platform-reported ACOS does not capture this.
Vendor vs Seller Central architecture. For brands selling on Amazon as Vendor (Amazon buys + resells) and Seller (the brand sells directly), the Amazon Advertising Account links to one or both. Vendor-side advertising (Sponsored Brands via Amazon's vendor product feed) and Seller-side advertising (Sponsored Products via the brand's seller listings) need coordinated account structure or the same product gets ads bidding against itself.
Brand-Registry-grade trademark + content controls. Brand Registry is mandatory for Sponsored Brands + A+ Content + Brand Storefront + Vine + Posts. Multi-marketplace brand registry requires a separate registration per marketplace (sometimes per regional brand office). Maintaining brand-registry hygiene per marketplace is the IP-protection + content-control prerequisite.
DSP + Sponsored Ads integration architecture
Sponsored Products + Sponsored Brands + Sponsored Display are the keyword + ASIN-targeted lower-funnel; Amazon DSP is the audience-targeted full-funnel + off-Amazon programmatic display + Streaming TV (STV) + audio. The integration matters because they share the customer's Amazon-attributed audience pool but spend out of different budget allocations.
The Dcrayons DSP + Sponsored Ads integration pattern:
Audience strategy spans both surfaces. Amazon's audience taxonomy (in-market, lifestyle, life-event, custom audiences from past purchase / search / view) is the unified audience layer. DSP campaigns target via Amazon's audience taxonomy; Sponsored Display Audience targeting uses the same. A coherent audience strategy applies the same audiences across both surfaces, with documented exclusion rules so the same shopper does not see ad-feed conflict.
Funnel placement coordination. DSP upper-funnel awareness ads + DSP STV + DSP off-Amazon display run the prospecting layer. Sponsored Brands + Sponsored Brands Video run the consideration layer. Sponsored Products + Sponsored Display Product Targeting run the consideration-plus-conversion layer. Sponsored Display Audience Retargeting + DSP cart-abandoner remarketing run the recovery layer. Each funnel stage has its own creative + budget + bid logic; the coordination is intentional, not accidental.
Retail-on-Retail discipline (RoaR). A frequent failure mode: Sponsored Display Product Targeting bids on competitor ASINs while DSP simultaneously bids on conquest audiences and Sponsored Products bids on the brand's own branded search. Three surfaces overlapping on the same shopper inflate spend without inflating outcome. The portfolio review enforces a Coordinated bidding rule: only one surface bids aggressively per shopper-segment per funnel-stage. The Amazon Marketing Cloud overlap report is the diagnostic.
Creative + brand-store coherence. DSP creative + Sponsored Brands creative + Brand Storefront creative + Posts creative + A+ Content all share the same brand-design language. Storyboard at the brand level; deploy across surfaces with format-specific renders. Creative drift across surfaces is the silent killer for brand consistency.
Amazon Marketing Cloud (AMC) measurement discipline
AMC is Amazon's clean-room measurement platform: ad-event + retail-event data joined at the pseudonymous user level, accessible via SQL queries, with results aggregated to prevent re-identification. AMC is the closest enterprise comes to ground-truth measurement on Amazon. The discipline is in knowing what to measure + what queries to write + how to interpret the results.
The Dcrayons AMC discipline:
Standard AMC report suite. Customer paths-to-purchase (which combinations of ad surfaces precede conversion), new-to-brand (NTB) percentage by surface, audience overlap across campaigns, frequency distribution + saturation, off-Amazon display contribution to on-Amazon conversion, STV uplift on attributable conversions. These are the standard quarterly board-deck queries.
Multi-touch path attribution. AMC's path-to-purchase queries reveal that a conversion typically involves 2-5 ad touches across Sponsored + DSP. The single-touch ACOS that the Amazon Ads UI reports under-credits upper-funnel surfaces. The AMC-aware allocation moves budget upward (more DSP STV + more Sponsored Brands) when the multi-touch data justifies it.
Incrementality testing. AMC supports geo-holdout + audience-holdout experiments. For a major brand or a flagship category, running a Q1 incrementality test (one Indian state held out from DSP retargeting; compare conversion rate vs the spent states) gives a defensible measurement of true incremental impact, separate from claimed attribution. Annual incrementality tests are how serious operators sanity-check Amazon's reported numbers.
Custom query library. Each enterprise account holds a documented library of AMC SQL queries that the team runs on a defined cadence: weekly path-to-purchase trend, monthly NTB-by-surface, quarterly incrementality, annual budget-allocation-optimisation. New queries are reviewed by a senior analyst + documented before being added to the cadence.
Closed-loop optimisation. AMC insights feed back into the campaign-management cadence. NTB-by-surface insights inform creative + budget shifts. Path-to-purchase insights inform funnel allocation. Audience overlap insights drive exclusion-list updates. AMC without a closed-loop integration to campaign management is an expensive academic exercise.
TACoS over ACOS: the cross-channel discipline
The most consequential governance rule on enterprise Amazon Ads: the metric to optimise is TACoS (Total Advertising Cost of Sale = Ad Spend / Total Brand Revenue on Amazon), not ACOS (Advertising Cost of Sale = Ad Spend / Ad-Attributed Revenue).
Why: ACOS is myopic. A brand with strong organic sales on Amazon can run high-ACOS Sponsored campaigns that drive incremental TOTAL revenue, including organic halo from the ad-driven discovery. Optimising to low ACOS pushes the team to defund those campaigns; total brand revenue falls; the team congratulates itself on the improved ACOS while the boss notices the revenue drop.
TACoS captures total brand-on-Amazon revenue against total ad spend. A healthy enterprise TACoS for an established brand is 5-12 percent; for a new launch brand 15-25 percent; for a brand in retention/defence posture 3-7 percent. The TACoS target is a CMO + CFO decision; the campaign-level ACOS is a tactical variable.
The Dcrayons TACoS discipline:
Weekly TACoS report at brand + marketplace level. Total ad spend across all surfaces (Sponsored + DSP) divided by total brand revenue on the marketplace (organic + paid). Trended weekly + against a quarterly target.
ACOS as a per-campaign tactical signal. Inside the campaign, ACOS still matters for the keyword + ASIN + bid-optimisation decisions. But the brand-level review uses TACoS.
Organic-revenue dependency awareness. A brand whose organic rank depends on continued advertising-driven sales-velocity needs to model the organic-revenue impact of ad-spend cuts. Cutting ad spend can cause organic rank decay, which causes organic revenue decay, which makes the new TACoS look worse even though ad spend went down. The MMM holds the elasticity; the campaign team operates within the MMM-defined envelope.
New-to-Brand (NTB) percentage tracking. Healthy programmes maintain NTB above 30 percent of ad-attributed conversions. Sub-30 percent suggests the programme is over-indexed on retargeting + branded-search defence and is not contributing to growth. The Sponsored Brands surface + DSP surface are where NTB is driven; budget tilt toward these surfaces when NTB drops.
Operational pattern past Rs 50 crore Amazon ARR
Enterprises that operate past Rs 50 crore Amazon-ARR (the threshold where ad-team-of-1 cannot keep up) apply these disciplines:
Catalog-quality SLA upstream. Listing-quality scores, A+ Content coverage, Brand Storefront freshness, Posts publishing cadence all have SLAs. Advertising performance is downstream of listing quality; a 90+ percent A+ Content coverage with 80+ percent listing-quality-score is the baseline.
Inventory-aware bidding. Sponsored Products bids ramp down when inventory drops below 21 days of supply; ramp to zero when out-of-stock. This is the operational guard that prevents spending against ASINs that cannot fulfil. Automated bid-rule, not human polling.
Event-readiness playbook. Prime Day, Diwali, Republic Day, GITEX, BFCM, Eid, Boxing Day each have documented prep playbooks: T-60 inventory plan, T-45 catalog refresh, T-30 creative + A+ Content updates, T-21 ad-bid lift schedule, T-14 day-of-event budget reservations, T-7 brand-store hero refresh, T-0 live monitoring.
Cross-marketplace coordination cadence. Weekly portfolio review across all marketplaces; monthly cross-marketplace strategy + budget review; quarterly governance + AMC deep-dive. Each marketplace's account team feeds into the central portfolio function.
Reporting cadence + dashboard suite. Daily: spend + revenue + TACoS per marketplace. Weekly: campaign-level review + creative-rotation + inventory-aware-bid-status. Monthly: NTB% trend + AMC path-to-purchase + incremental-margin review. Quarterly: AMC incrementality + cross-marketplace MMM-driven budget reallocation.
Production checklist: the rollout sequence
For an enterprise Amazon Ads programme at Rs 50+ crore Amazon ARR across 3+ marketplaces:
- Multi-marketplace account topology mapped: Advertising Account per marketplace, brand-Asset-Group organisation, billing + tax per marketplace
- Brand Registry maintained in every marketplace, A+ Content + Brand Storefront live per marketplace
- PIM-driven catalog mapping per marketplace, ASIN-to-master-product traceability
- Vendor vs Seller advertising coordination if both account types present
- Sponsored Ads campaign structure: per-funnel-stage + per-brand-line, never cross-brand
- DSP integration: audience strategy shared with Sponsored Ads, funnel-stage coordinated bidding
- AMC instance provisioned, standard query suite + custom query library documented
- AMC closed-loop cadence: weekly query + monthly review + quarterly incrementality test
- TACoS target locked + agreed with CMO + CFO, weekly TACoS report at brand + marketplace level
- Inventory-aware bidding automation, out-of-stock guard
- Event-readiness playbooks per marketplace per major event
- NTB% tracking + Sponsored Brands + DSP allocation discipline
- Brand-safety + brand-registry monitoring + counterfeit takedown workflow
- Reporting cadence: daily / weekly / monthly / quarterly with named owners per cadence
References + linked context
- Amazon Ads documentation: Sponsored Products, Sponsored Brands, Sponsored Display, Amazon DSP, Amazon Marketing Cloud
- Dcrayons glossary: tacos, acos, amc, ntb, sponsored-brands, dsp
- Dcrayons Amazon SPN reference architecture: see /learn?tag=amazon-spn for the catalog + Brand Registry + Vendor-vs-Seller operational pattern Amazon Ads pairs with
If your enterprise Amazon Ads programme is fighting multi-marketplace portfolio governance, DSP + Sponsored coordination, or the move from ACOS to TACoS discipline, this is the architecture we deploy. Reach out via the contact form for a 30-minute review against your current setup.
