Three Amazon Advertising developments from Q1 2026 deserve attention from brands and developers building on the Ads API. None received much press coverage, but each has meaningful implications for how Amazon advertising data is accessed, managed, and attributed.
1. Brand Stores Management API: Now Generally Available
Amazon’s Brand Stores Management API exited beta and reached general availability in February 2026. This makes programmatic management of Amazon Brand Stores production-ready for the first time.
What the Brand Stores API Provides
For brands unfamiliar: Amazon Brand Stores are multi-page branded storefronts within Amazon — essentially a mini-website inside Amazon’s ecosystem. They can include product galleries, lifestyle images, videos, and curated collections. Access requires Brand Registry.
The Brand Stores Management API provides programmatic access to:
Store structure management:
- Create, update, and delete store pages
- Manage page hierarchy and navigation
- Add and remove content modules (product tiles, text, video, image tiles)
- Schedule seasonal updates without manual Store Builder sessions
Section-level analytics (new in GA): The most significant new capability. You can now query performance metrics at the individual store section level:
{
"metrics": [
"renders",
"viewableImpressions",
"clicks",
"clickThroughRate",
"sales14d",
"unitsSold14d"
],
"groupBy": ["storePageId", "moduleName"],
"dateRange": {
"startDate": "2026-01-01",
"endDate": "2026-02-28"
}
}
This returns renders, viewableImpressions, clicks, and clickThroughRate per storefront section — telling you which tiles, modules, and pages within your Brand Store are performing, not just aggregate store traffic.
Why this matters: Brand Stores previously offered only top-level metrics in Seller Central — total visits, total sales attributed. You couldn’t tell whether users who entered via the “New Arrivals” page converted better than those who entered via the “Best Sellers” page. Section-level metrics enable genuine storefront optimisation rather than intuition-based design decisions.
Practical Use Cases
Automated seasonal updates: Use the API to push seasonal content changes to Brand Store pages without manual editing in Store Builder. A Christmas product showcase can be scheduled and activated via an n8n workflow that fires on December 1 and reverts on January 6.
Performance-based layout optimisation: Pull section metrics weekly, identify lowest-performing modules, and automatically move them or flag them for creative refresh. High-performing sections can be promoted to more prominent positions.
Multi-brand management: Agencies managing Brand Stores for multiple clients can use the API to push consistent updates (price changes, seasonal messaging, compliance updates) across all stores programmatically.
Reference: Brand Stores Management API documentation
2. Amazon Ads MCP Server: AI Agents for Campaign Management
In February 2026, Amazon opened the Amazon Ads MCP Server to open beta. MCP (Model Context Protocol) is a standard developed by Anthropic that lets AI assistants (Claude, ChatGPT with plugins, etc.) interact with external services through a structured tool interface.
What This Means in Practice
The Amazon Ads MCP Server exposes Amazon Advertising API capabilities as MCP tools. An AI assistant with the MCP server connected can:
- Answer questions about your Amazon Ads campaigns in natural language
- Pull performance reports without writing API code
- Suggest bid adjustments based on performance data
- Draft campaign briefs from historical performance patterns
Example workflow:
User: “What are my top 5 best-performing Sponsored Products campaigns by ROAS for the last 30 days?”
The AI assistant:
- Calls the MCP tool
get_campaignsto list active campaigns - Calls
get_campaign_metricswith date range and ROAS metric - Sorts and returns the top 5 with current bids, impressions, spend, and ROAS
- Presents the result in natural language with a table
No code required. No API call construction. No authentication flow to navigate.
User: “For the lowest ROAS campaign, what keywords are dragging it down?”
The AI assistant continues the conversation, drilling into keyword-level data for that campaign.
How to Access the Beta
The Amazon Ads MCP Server is currently in open beta. To access:
- Go to advertising.amazon.com/API
- Look for the “MCP Server” section under developer tools
- Follow the registration process for beta access
Once registered, you receive an MCP server endpoint and authentication token that you configure in your AI assistant’s settings.
Current limitations in beta:
- Read-only (you can query data but not create/modify campaigns via MCP yet)
- Response latency is higher than direct API calls (the MCP middleware adds overhead)
- Supported AI assistants: Claude Desktop (via Anthropic’s MCP host), and any MCP-compatible client
Write access (creating campaigns, updating bids, adding keywords) is planned for a future release.
Why This Matters Beyond the Convenience
The MCP Server represents a pattern shift in how advertising APIs will be consumed. Instead of developers writing custom code to query and display Amazon Ads data, AI assistants with MCP connections become the interface layer.
For brands and agencies who want Amazon Ads insights without developer resources, this is a significant accessibility improvement. For developers, it’s a new integration pattern to build services around.
3. New Shopping-Signal Enhanced Attribution Model (From January 1, 2026)
The least-publicised of the three changes, but potentially the most impactful for how you interpret performance data.
What Changed
From January 1, 2026, Amazon updated the attribution model for Amazon Store Ads (also called Brand Store ads or Sponsored Brands video) to incorporate “shopping signal enhanced last-touch attribution.”
The change specifically affects how view-through conversions are counted and credited for Store Ad placements.
The Old Model vs. The New Model
Old model: View-through conversions for Store Ads used Amazon’s standard view-through attribution — if a shopper viewed your ad but didn’t click, then later purchased the advertised product within the attribution window, the conversion was attributed to the view-through.
New model: The shopping-signal enhanced model applies additional filters before attributing a view-through conversion. It looks at Amazon’s first-party shopping signals — search behaviour, product page views, purchase history — to assess whether the ad view was likely to have actually influenced the purchase.
In plain terms: the new model is more conservative in claiming view-through credit. It discounts view-through conversions where the purchase behaviour suggests the customer would have bought regardless of seeing the ad.
Impact on Your Performance Data
You may see a decrease in reported conversions and ROAS for Store Ads campaigns — not because performance actually declined, but because the attribution model is now more conservative about which conversions it claims.
This is actually a more accurate picture of your campaign’s incremental impact. The old model over-attributed conversions to ad views; the new model attempts to isolate genuinely incremental conversions.
If you use ROAS as a campaign health metric, establish a new baseline from January 1, 2026 data before comparing to pre-January performance. Comparing December 2025 ROAS to February 2026 ROAS is an apples-to-oranges comparison if Store Ads are a significant part of your mix.
For DSP campaigns: This attribution change applies to Store Ads specifically. Standard Sponsored Products and Sponsored Brands campaigns use separate attribution windows configured at the campaign level and were not changed in this update.
Connecting These Changes to Your 2026 Strategy
Brand Stores GA → automate storefront operations: If you manage a large Brand Store with frequent product updates, the API now lets you systematise this. A weekly n8n workflow that pulls section performance metrics and flags underperforming tiles is now straightforward to build.
MCP Server → access your data without code: For teams that want Amazon Ads insights but don’t have dedicated analytics developers, the MCP Server beta is worth registering for. Early access while it’s free and in beta is low-risk.
Attribution model change → recalibrate benchmarks: If your reporting shows a January dip in Store Ads performance, check whether the attribution change explains it before optimising campaign settings based on what may be a measurement artefact.
We work with Amazon sellers and agencies on SP-API integrations, Advertising API automation, and multi-channel reporting. If you need help building on any of these new capabilities, book a free consultation.