Adobe Commerce Optimizer Implementation
A faster storefront and better product discovery, without ripping out the backend your business runs on. We connect your catalog to ACO, model it properly, and get search, recommendations and the storefront live.
What is Adobe Commerce Optimizer?
Adobe Commerce Optimizer (ACO) is Adobe's SaaS layer for the front of your store. It sits on top of an existing commerce backend and adds a flexible catalog model, AI-driven search and recommendations, merchandising tools, and a high-performance storefront on Edge Delivery. Your current platform keeps cart, checkout and orders.
Who ACO is built for, and who should skip it
Adobe positions ACO for three kinds of businesses:
Merchants who want to keep their backend and change only the storefront experience
Businesses where another system already runs cart and checkout
Adobe Experience Manager customers who want a simpler way to manage product data coming from a third-party commerce engine
Where we'd tell you to wait.
If your catalog is small, you sell on one site with one price list, and your storefront already passes Core Web Vitals, ACO's biggest benefits (catalog views, price books, the Edge storefront) won't earn back the licence quickly. We'd rather say that on the first call than after a six-figure project.
How Adobe Commerce Optimizer works, in plain terms
Think of it as four layers:
01
Your source systems.
Adobe Commerce, a PIM, an ERP, or another commerce platform. They stay the system of record.
02
Catalog ingestion.
Product, attribute, price and category data flows into ACO's Merchandising Services.
03
Catalog views and policies.
You decide who sees which products at which prices: a dealer network, a regional brand, a marketplace feed, a B2B account group.
04
Delivery.
Storefronts query ACO through GraphQL for search, listings and recommendations. Cart and checkout calls go back to your commerce platform.
For Adobe Commerce stores on Adobe's cloud infrastructure or self-hosted, Adobe provides a first-party Adobe Commerce Optimizer Connector. For other backends, data goes in through the Catalog Data Ingestion API, and cart handoff is built with API Mesh and App Builder.
Which path applies to you?
| Your current setup | How ACO connects | What we build |
|---|---|---|
| Adobe Commerce 2.4.7+ (cloud or on-premises) | First-party ACO Connector (Composer package) | Connector setup, scope mapping, catalog views, storefront |
| Adobe Commerce older than 2.4.7 | Upgrade first; the connector requires 2.4.7+ | Upgrade, then connector (upgrade service) |
| Another commerce engine or ERP-led stack | Catalog Data Ingestion API | Custom ingestion feed, API Mesh + App Builder for cart handoff |
| Planning a move to Adobe Commerce as a Cloud Service | Connector as a phased stepping stone | Staged roadmap |
What Evrig does on an ACO project
Readiness audit.
We check your Commerce version, installed Adobe SaaS extensions, catalog size, store view and website structure, and price complexity against ACO's licence limits before anyone signs anything.
Connector installation and configuration.
Composer install of adobe-commerce/commerce-data-export-aco-adapter, IMS OAuth server-to-server credentials, scope export settings, and the aco:config:init setup, all done on staging first.
Catalog modelling.
We design catalog sources, price books, catalog views and policies around how you actually sell, not just around how your store views happen to be set up today.
Search and merchandising setup.
Facets, synonyms, merchandising rules and recommendation units, configured in ACO Studio.
Storefront delivery.
A Commerce Storefront on Edge Delivery Services, or a custom headless front end that points at the correct ACO tenant and catalog view.
Cart and checkout handoff.
Where checkout lives in another platform, we build the handoff with API Mesh and App Builder so order flow is untouched.
Eventing and measurement.
We verify storefront events in ACO's Events page (recommendations depend on them) and set a before-and-after KPI baseline so you can prove results.
What we check before switching the connector on
These are the details that cause most failed first syncs. They come straight from Adobe's setup requirements and limits.
01
Conflicting extensions come out first.
Live Search, Product Recommendations, Catalog Service and the Data Management Dashboard must be uninstalled before the connector is enabled. Leave them in and Adobe warns you'll get broken configuration screens, duplicate data in ACO, and 401/403 authentication errors. Your existing Live Search and Recommendations settings in the Commerce Admin also stop applying, so we export and document your current rules first.
02
Environments must match.
Sandbox ACO instances connect to non-production Commerce environments; production connects to production. Mixing them gives inconsistent catalog data, search results and recommendations. Each ACO instance includes two sandboxes and one production environment.
03
Scopes are decided before the first sync.
Each store view becomes a catalog source. Each website and customer group combination becomes a price book. Changing export settings later triggers a full re-index, which on a large catalog can take hours. We lock the scope plan first.
04
Disabling a store view deletes its storefront data.
The catalog source stays, but synced data is removed on the next cron run. Teams find this out the hard way when "just turning off" a test store view.
05
The first sync is not instant.
Adobe says it can take from a few minutes to several hours depending on catalog size. We schedule it and check results in both Data Feed Sync Status (Commerce Admin) and Data Sync (ACO).
06
The connector is built for B2C.
Adobe documents the connector's supported scenarios for B2C merchants. If you run B2B features, we scope that gap on the readiness call.
The limits worth knowing before you sign
Adobe's marketing headline says the catalog model can handle up to 250 million SKUs per instance and up to 30,000 price points per SKU. That describes what the architecture can scale to. Your base licence is much smaller, and you pay to expand it. Size your contract on these numbers:
| Area | Base allocation | Can it grow? |
|---|---|---|
| SKUs per catalog source | 250K | Yes, in 100K packs |
| Catalog variations (catalog views × price books) | 100 | Yes, in 100-variation packs |
| Product updates | 1K per minute, 100K per day | Yes, up to a 1M/day ceiling |
| Product payload size | 200 KB | No |
| Variants per product | 10K | No |
| Catalog sources | 50 | No |
| Policies per catalog view | 10 | No |
| SKUs per GraphQL product request | 100 | No |
| Search pagination depth | 10K products | No |
| Active recommendation units | 50 | No |
Two things we flag on almost every project:
Catalog variations multiply.
Five catalog views with 20 price books is already 100 variations. Stores with many customer groups can hit the base limit before launch.
Recommendations are simpler than in Adobe Commerce.
Category or attribute inclusions and exclusions in recommendation units are not supported in ACO. If your current recommendations depend on those filters, we plan an alternative before migration.
Generative AI image creation needs a separate Firefly licence on the same IMS org, and AI credits are bought separately. Budget for them if the AI features are part of the business case.
What drives the cost of an ACO project
We don't quote a flat price because the work depends on your setup. These are the factors that move the number:
- Adobe licence (paid to Adobe, separate from our fee), plus any SKU, variation or ingestion packs
- Your backend: Adobe Commerce with the connector is the fastest route; a non-Adobe backend needs a custom ingestion feed
- Catalog model complexity: number of brands, regions, customer groups and price books
- Storefront choice: Adobe's prebuilt Edge Delivery storefront blocks versus a custom headless build
- Checkout handoff: none, or a custom API Mesh / App Builder integration
- Existing search rules: how much Live Search configuration needs rebuilding in ACO
Common concerns
"Is this a replatform in disguise?"
No. With the connector, Adobe Commerce stays the system of record for products, prices and customer groups. The connector doesn't modify cart, checkout or order flows.
"Are we locked in?"
Your catalog data still lives in your backend. ACO holds a synced copy. We hand over all code, App Builder projects and documentation.
"What if the sync breaks?"
Feed health, last sync times and per-SKU status are visible in Data Feed Sync Status. Targeted resyncs can be run per feed without a full reload.
FAQ
Connect ExpertFind out if ACO fits before you buy the licence
Send us your Commerce version, rough SKU count and number of sites. We'll tell you which path applies, where the licence limits bite, and what the project involves.