Which open source ecommerce platform fits your business
If you want to own your storefront, customer data and hosting, the right open source ecommerce platform depends less on features and more on three things: how complex your catalog and pricing rules are, how heavily the storefront must load your ERP and order systems, and how much in-house engineering capacity you can keep staffed after launch. For most organizations the first split is between a monolith that bundles storefront, catalog and checkout in one application, and a headless commerce API that serves those functions to separate front ends. Magento Open Source (OSL-3.0, with 2.4.9 as the latest stable tag in its repository) and WooCommerce (GPL version 2 or later, with 11.2.1 published on October 9, 2026) are the monolith families most often considered, while Saleor (BSD-3-Clause), Medusa (MIT, except materials under its Enterprise Edition license) and Sylius (MIT, with v2.3.1 published on October 14, 2026) are headless options that suit teams building their own storefront.
A monolith usually reduces the number of moving parts your team must patch and test, which matters when the store is run by a small group. WooCommerce inherits the WordPress hosting model and plugin ecosystem, which speeds up simple catalogs but can grow fragile when many extensions interact. Magento Open Source carries more built-in catalog, promotion and B2B-style pricing depth, but it also expects more hosting capacity and specialist skills. Headless platforms such as Saleor, Medusa and Sylius push more design and integration work to your engineers, and they reward that effort when channels, marketplaces or custom checkout flows multiply. Before shortlisting, ask each vendor and implementation partner to demonstrate your three hardest cases, such as multi-warehouse inventory, contract pricing and tax rules, against a working build rather than a slide.
Confirm the license terms with your legal team before you commit, because a license can change what you may host, modify or resell, and some vendors keep enterprise features in separate editions. Also confirm the current release line and its support window directly with the project, since the researched facts in this section list only version numbers and dates, not support lengths. Finally, decide who will own patching, payment integration and ERP connectors after go-live. That operating decision, more than the platform name, determines whether the store stays secure and current for the next several years.
Requirements to agree before choosing an open source ecommerce platform
Platform choice follows from a short set of business facts, not from feature lists. The decisions below shape whether a monolith or headless architecture fits, which integrations need prebuilt connectors, and who carries the operating load after launch. Agree them in writing before comparing Magento Open Source, WooCommerce, Saleor, Medusa or Sylius.
Catalog and pricing complexity
Count SKUs, variants, bundles, configurable products, B2B price lists, customer-specific contracts and promotion rules. A catalog of a few hundred simple products fits very different platforms than one with tens of thousands of variants and negotiated prices. Ask each candidate how it stores attributes, indexes product data and handles price rules at your expected volume, and which of those behaviors come from core versus extensions you would maintain.
Multi-store and multi-currency needs
Decide how many storefronts, brands, countries, languages, tax regimes and currencies you must run from one codebase and one order flow. Each added storefront multiplies content, catalog rules, tax configuration and payment setup. Confirm whether the platform supports shared catalogs with per-market pricing, and whether currency conversion and tax calculation are built in or handled by a third-party service you would need to contract.
ERP and order management integration
Identify the system of record for inventory, pricing, invoicing and fulfillment, and the direction and frequency of each data flow. Real-time stock sync, returns handling and split shipments require more integration work than nightly order exports. Ask which ERP connectors exist for your system, whether they are maintained by the platform community or a vendor, and who owns the middleware that maps fields between them.
Checkout and payment methods
List the payment methods, wallets, buy-now-pay-later options and regional schemes you must accept, and whether card data will touch your servers at all. Hosted or tokenized checkout keeps most card data outside your environment and narrows PCI DSS scope. Direct card entry on your own pages expands it significantly. Confirm which payment gateways have maintained plugins for each candidate and how refunds, partial captures and chargebacks are handled.
Peak traffic and seasonal load
Estimate expected peak concurrent shoppers, promotion-driven spikes and the number of product searches and API calls at those peaks, and measure them against current analytics rather than hope. Peaks drive hosting architecture, caching layers, search infrastructure and the load testing you must budget for before a sale event. Ask each vendor or partner for reference architectures at your traffic level and for the results of load tests they have run on comparable catalogs.
Release and security patch ownership
Decide who applies core upgrades, extension updates and security patches, on what cadence, and who signs off before production. Open-source commerce platforms are not updated by the vendor on your behalf. Confirm which release lines are supported and for how long, which extensions you depend on and whether they are still maintained, and whether an internal team or a retained partner will own the patch cycle, testing and rollback plan.
Magento Open Source, WooCommerce, Saleor, Medusa and Sylius compared
This table compares the five named platforms by license, latest supplied release, architecture and the operating effort each implies. Read the license and architecture rows first, then the fit row, because they set the integration and hosting work you will fund. Licenses describe the core only, so confirm the terms of every extension, theme and module you plan to use.
| Capability | Platform | Trade-off | When to choose it |
|---|---|---|---|
| Magento Open Source | Licensed under OSL-3.0. Latest stable tag 2.4.9 in the magento2 repository. | Monolithic storefront and back end in one application, with heavier hosting and tuning needs. | Large catalogs and complex B2B or multi-store catalog rules, with in-house or partner engineering capacity. |
| WooCommerce | Licensed under GPL version 2 or later. Version 11.2.1 published October 9, 2026. | Runs on WordPress, so content and commerce share one stack and plugin dependencies grow over time. | Content-led storefronts and smaller catalogs where the marketing team already works in WordPress. |
| Saleor | Headless commerce API licensed under BSD-3-Clause. Version 3.23.42 published October 16, 2026. | API-first design means you build and operate the storefront and integrations yourself. | Teams with frontend engineering capacity that want commerce logic decoupled from presentation. |
| Medusa | MIT licensed, except materials covered by its Enterprise Edition license. Version 2.21.2 published September 28, 2026. | Modular, headless architecture. Confirm which features sit in Enterprise Edition materials before you rely on them. | Developer-led teams building custom commerce flows on a Node.js stack with a separate storefront. |
| Sylius | Licensed under MIT. Version 2.3.1 published October 14, 2026. | Symfony-based framework that suits custom domain logic, but requires PHP engineering skills to maintain. | Organizations with PHP teams that need deep customization of catalog, pricing and checkout. |
| Architecture style | Monolith (Magento, WooCommerce) versus headless API (Saleor, Medusa, Sylius) | Monoliths ship faster out of the box. Headless designs separate storefront releases from commerce releases, adding integration and hosting work. | Choose headless when multiple channels or a custom storefront are firm requirements, not just preferences. |
| Operating effort | Recurring upgrades, security patching, hosting tuning and extension maintenance across all five | Effort depends on customization, traffic peaks and the number of extensions, so ask partners to estimate it from your written scope. | Plan for ongoing patch cycles and a named owner for each release, regardless of license. |
| PrestaShop (alternative) | Core licensed under OSL-3.0. Modules licensed under AFL-3.0. Version 9.2.0 published September 30, 2026. | Monolithic storefront with a module marketplace, so module licensing must be reviewed separately. | Smaller merchants in regions where PrestaShop has an established partner and module ecosystem. |
When a monolith is enough and when headless pays off
A monolith keeps the storefront, catalog, checkout, admin and extensions in one codebase deployed as one application. For a modest product range sold through a single website, this is usually the shortest path to launch. WooCommerce, which runs on WordPress, and Magento Open Source, which ships with its own storefront, both follow this model, and PrestaShop does too. Your team works in one place, upgrades move together, and most themes and extensions are built for this shape.
- Fewer moving parts: one application, one database and one deployment pipeline to operate and back up.
- Lower frontend staffing: the storefront comes with the platform, so a smaller team can ship without a dedicated JavaScript build chain.
- Coupled releases: a storefront change and a checkout change ship together, which simplifies testing but widens the blast radius of each upgrade.
- Extension dependence: most business logic lives in plugins or modules, so review their maintenance history before committing.
- Hosting profile: a single application tier with a managed database is often enough, though load requirements must be confirmed against your own traffic.
Before choosing this path, confirm with the vendor which release line is supported and how long security fixes are guaranteed for it.
Headless commerce separates the commerce engine, exposed through APIs, from one or more storefronts that consume those APIs. Saleor and Medusa are API-first, and Sylius is built around a service layer that other applications call. This model suits organizations that sell through a web app, a mobile app, a marketplace feed, a point-of-sale system or a partner portal from the same product data and pricing rules. It also suits brands that need a storefront design the platform's default theme cannot deliver.
- Integration effort rises: every channel needs its own API integration, authentication handling and error paths.
- Independent release cycles: the frontend can deploy several times a week while the commerce core upgrades on a slower schedule.
- Frontend skills are required: plan for a team that owns the storefront framework, rendering strategy, caching and accessibility testing.
- Hosting becomes plural: the API, storefront, search index, media delivery and background workers often scale separately.
- Accessibility is your responsibility: a custom storefront needs its own testing against WCAG 2.1 Level AA, which public-sector buyers may require.
Ask any implementation partner to show how the API contract is versioned and how each channel is tested against it before go-live.
Many organizations already run a content management system for editorial pages, campaigns and help content, often Drupal, WordPress or a headless CMS such as Strapi, Payload or Directus. In that case the question is less about the commerce engine and more about where product content, landing pages and checkout experiences should live. A decision here determines who edits what, how previews work and how the two systems stay consistent.
- Keep the CMS for content: editorial pages, guides and campaign pages can stay in the existing system while commerce handles products, carts and orders.
- Define data ownership: decide which system is the source of truth for product descriptions, media, pricing and promotions to avoid duplicate records.
- Preview across two systems: editors need a way to see a campaign page with live product data, which adds integration work.
- Avoid double licensing and hosting: confirm the license terms of each component, since some open-source CMS products separate community and enterprise features.
- Plan a shared identity model: customers, staff and partner accounts should authenticate once across both systems where possible.
Review the existing CMS's release and support window before wiring it to commerce, so that two upgrade calendars do not collide.
Reference architecture components and how they connect
A self-hosted commerce stack is a set of components that each own one job: presenting the catalog, holding prices and orders, finding products, moving data to the ERP, taking payment and keeping records safe. The decisions that matter most are where requests cross between these components, which of them can fail without stopping checkout, and which parts sit inside your own hosting and high availability design.
Storefront layer
The storefront renders pages, carts and checkout for shoppers. In a monolith, it is the same application as the commerce core. In a headless design, it is a separate front-end application that calls the commerce API, so you own its hosting, caching and release pipeline. Check whether product pages can be cached at the edge, how cart state is held between requests, and who patches the front-end framework. Confirm which Magento, WooCommerce, Saleor, Medusa or Sylius option ships a storefront and which one you must build.
Commerce core and order service
The commerce core owns the catalog, pricing, tax rules, inventory reservations, carts and order state. It is the system of record that the storefront, ERP and payment layer all depend on. Decide early whether it runs as a single application with a database or as separate services for catalog and orders, because this sets your upgrade path and your failure domain. Ask the implementation partner how the core is versioned and how long your chosen release line receives security fixes.
Search and catalog index
Shoppers rarely browse a catalog of any size without search, facets and filters. A dedicated search engine indexes products from the commerce core and serves queries so the core database is not loaded by every keystroke. The index is rebuilt or updated from change events, so stale stock or price data can appear in results. Confirm the indexing latency you need, how reindexing runs during peak trading, and whether the search cluster is inside the same high availability zone as the storefront.
Integration middleware and queue
Orders, stock levels, customer records and invoices move between commerce and the ERP, tax engine and warehouse. A message queue or integration service between them absorbs spikes, retries failed calls and keeps a record of each message. Without it, an ERP outage can block checkout or lose orders. Check where the queue is hosted, how dead-letter messages are reviewed, how idempotent order writes are, and who owns the mapping between commerce fields and ERP fields.
Payment layer and tokenization
The payment layer connects the storefront to a payment processor. The strongest scope reduction comes from hosted fields or redirects, where card data is entered on the processor's pages and your servers only receive tokens. Your PCI DSS scope then depends on how the storefront, templates and scripts are built, not only on the processor. Confirm which integration pattern your design uses, who manages payment plugins and script changes on checkout pages, and how refunds and chargebacks reach the order service.
Data stores, backups and hosting
The database, media storage, session store and search index each need their own backup schedule, tested restore procedure and retention rule. Restoring the database without matching media files or search indexes leaves orders pointing at missing assets. Decide where replicas sit, whether the database fails over automatically between zones, and how the stack is rebuilt in a second region. Ask the partner to show a restore test that covers the full stack, not only the database dump.
Build in-house, work with a partner or buy SaaS for ecommerce
This table compares building on open-source ecommerce in-house, running it with an implementation partner, and buying a commercial SaaS platform. Read each row against your own team, data residency rules and cost profile, because the answer usually depends on who carries patching, PCI scope and upgrades. The main takeaway is that open source removes license fees but moves recurring operating work onto someone, and that someone must be named in the contract.
| Factor | In-house open source | Open source with a partner | Commercial SaaS |
|---|---|---|---|
| Control over code and data | Full control of code, hosting location and data flows | Full control retained in your cloud or data center, with the partner working inside it | Code and platform roadmap set by the vendor, data hosted in vendor-selected regions |
| Internal skills needed | Platform, security, PCI and ecommerce engineering on staff year-round | A smaller internal product owner and integration lead, with specialist delivery supplied under contract | Storefront and merchandising skills, plus vendor administration |
| Who pays for hosting | Your organization, through cloud or data center accounts you run | Your organization for infrastructure, with the partner billed for implementation and support scope | Usually included in subscription fees, confirm what is metered |
| Who patches and upgrades | Your team, on the release cadence you set | Shared by written scope, confirm release windows and emergency patch response | Vendor, on the vendor's schedule |
| Core cost structure | Mostly recurring staff and infrastructure, with one-time build effort | One-time implementation plus recurring retainer or support, plus infrastructure | Recurring subscription, often tied to order volume or GMV, plus one-time migration and integration |
| Who owns PCI DSS evidence | Your organization, including segmentation and the self-assessment or assessment | Your organization remains accountable, the partner prepares configuration and evidence | Shared responsibility matrix from the vendor, confirm what they will attest to |
| Main cost risk | Under-staffed operations after launch | Unclear scope for upgrades, testing and incident response | Price growth as volume increases, and limits on custom integrations |
| Best fit | Large engineering teams with an existing platform function | Regulated or data-sensitive businesses that want open code without running every operation themselves | Teams that want fast launch and accept vendor-set roadmap and pricing |
How an ecommerce rollout runs from discovery to hypercare
The sequence turns a commerce ambition into a platform that takes payments, syncs with the ERP, and runs under an agreed support model. Each phase closes with a written artifact so the business, security and finance owners can sign off before the next phase begins.
Run discovery and capture requirements
DeployOpen works with merchandising, operations, finance, security and the current commerce team to document the catalog structure, pricing and promotion rules, tax and shipping logic, B2B needs, peak traffic expectations and the systems that must connect (ERP, tax, OMS, search, CRM). The output is a requirements register with each item marked as must have, should have or later, plus a list of open questions. The phase is done when the business owners have signed the register and the open questions have named owners and due dates.
Select the platform and design the architecture
Using the requirements register, the team compares Magento Open Source, WooCommerce, Saleor, Medusa and Sylius against the business model, the team's skills and the integration load, and decides between a monolith and a headless storefront. The output is a short architecture decision record covering hosting target, environments, the storefront approach, the search and caching layers, and the integration pattern. It is complete when the architecture has been reviewed by the security team and the platform owner has accepted the support model in writing.
Build the storefront and integrations
Engineers configure or customize the chosen platform, build the storefront, and implement integrations with the ERP, tax calculation, shipping carriers, identity provider and marketing tools. Each integration is tested against agreed sample data in a non-production environment, with error handling and retry behavior documented. The output is a tested build with an integration catalog listing each interface, its data owner, its failure mode and its monitoring. The phase closes when the integration test cases pass and the open defects are triaged.
Scope payments and PCI obligations
Before the checkout is finalized, DeployOpen and the merchant's payment team map which components touch cardholder data, whether hosted payment fields or redirects keep card data out of the merchant's servers, and which requirements of the PCI DSS apply to the chosen design. The output is a written scope statement with a data flow diagram and a responsibility matrix that separates the merchant's duties from the payment provider's. The phase is complete when the acquirer or qualified assessor has reviewed the scope and the team has confirmed it in writing.
Migrate data and rehearse cutover
The team extracts products, customers, orders, redirects and tax data from the source system, transforms them to the new schema, and loads them into staging. Reconciliation checks compare record counts, sample orders and totals against the source. The cutover runbook is then rehearsed at least once end to end, timing each step and recording rollback triggers. The output is a reconciliation report and a signed runbook. The rehearsal is done when the team can run the cutover from the runbook without improvising a step.
Go live and run hypercare
The team executes the cutover during the agreed window, validates live transactions, DNS, certificates, redirects and integrations, and confirms monitoring and alerting are receiving data. Hypercare follows with an elevated support rota, daily defect review and a watch on payment failures, order sync errors and search or cache performance. The output is a go-live report and a handover pack with runbooks, access records and the support model. Hypercare ends when the agreed exit criteria are met and the platform owner accepts the handover in writing.
Migrating from Shopify or Salesforce Commerce to open source
Start the migration plan with a full inventory of what the source platform can export, because that list, not the platform's marketing, sets the migration scope. For Shopify and Salesforce Commerce, confirm in the source environment which objects come out through native exports, APIs or bulk tools: product catalog with variants, categories and attributes, customer accounts and password handling, order history, gift cards, coupons, product reviews, CMS pages and media. Password hashes are the usual blocker. Many platforms will not export them in a usable form, so plan for a forced password reset or a staged account claim flow. Record every object as moved, transformed or left behind, and get that decision signed off by the business owner for orders and customers, since those carry tax, legal and support obligations.
Redirects carry the most SEO risk. Build a URL map from every indexed product, category, filter and content page on the old site to its new path, then implement it as server-side 301 redirects before the cutover, not after. Check that filtered and paginated URLs with search value are either mapped or deliberately removed, and that canonical tags, hreflang and the XML sitemap point to the new structure on launch day. Expect rankings and crawl activity to shift for a period after any URL change, and ask the implementation partner how redirect coverage will be tested against the crawl data from the live store.
What must be rebuilt is usually more expensive than what is exported. Shopify apps, Salesforce Commerce cartridges and their storefront code do not transfer to Magento Open Source, WooCommerce, Saleor, Medusa or Sylius. List every app, custom checkout step, tax or shipping rule, pricing engine call and ERP, tax, fraud or marketing connector, then decide for each whether to replace it with an open-source module, rebuild it in the new codebase or drop it. Payment tokens generally cannot be moved between gateways, so plan a reconnection of stored payment methods with each provider, and confirm the card data path with your assessor before the checkout is rebuilt, since it determines your PCI DSS scope.
Systems an open source ecommerce platform must connect to
A storefront is only as useful as the systems behind it. Orders, stock, tax, payments, customer records, and sign-in all have to agree, and each connection carries its own data ownership and failure rules. The cards below cover the integrations that most often shape the platform choice and the implementation budget.
ERP and inventory sync
The ERP usually owns product master data, stock levels, pricing and order financials, so the storefront must read and write against that source of truth. Decide which system owns each field, how often stock is pushed, and what happens when an order is placed against stock that the ERP has already allocated. Ask the ERP vendor which APIs or connectors it exposes for orders, inventory and price lists, and whether they handle bulk or only event-level updates. Plan for reconciliation jobs, because webhooks alone do not catch missed messages.
Tax calculation
Sales tax and VAT rules change by jurisdiction, product category and customer status, so most teams connect a tax engine rather than maintain rate tables in the platform. Confirm which tax engine the platform's extensions support, how exemption certificates are stored, and whether invoices carry the tax lines your accounting system requires. Ask who maintains rate updates and how quickly changes reach production. Test with real destination scenarios before go-live, including cross-border orders and B2B exemptions.
Shipping and fulfillment
Shipping rates, carrier labels, tracking events and warehouse handoffs all sit between checkout and the customer's delivery. Decide whether rates are calculated in the storefront or pulled from a carrier or fulfillment system, and who owns the shipment status once an order leaves the warehouse. Confirm which carriers have maintained connectors for your platform version, and whether tracking updates return to the order record. Missing status updates are a common source of support volume after launch.
Payment gateways and PCI scope
Each payment connection determines how much cardholder data your infrastructure touches. Hosted fields or redirect flows keep card data off your servers and usually shrink PCI DSS scope, while self-hosted card capture expands it substantially. Confirm with your acquirer and gateway which integration model they support, whether 3-D Secure is handled, and how refunds and chargebacks post back to the order and ledger. Ask the gateway for its attestation documents and responsibility matrix before you design the network.
CRM and marketing automation
Customer, lead, consent and campaign data must move between the storefront and the CRM without creating duplicate contacts. Define the system of record for customer identity, how guest checkouts merge into existing profiles, and which events (cart abandonment, order placed, refund issued) trigger marketing workflows. Confirm the consent fields that travel with each contact, and whether opt-out status is enforced before any email or SMS is sent. Ask the CRM vendor for its supported sync direction and rate limits.
B2B accounts, identity and single sign-on
Wholesale and B2B buyers often need company accounts, approval hierarchies, negotiated prices and purchase on account, which many storefronts support only through add-ons or custom code. Connect the platform to your identity provider over OpenID Connect or SAML so staff sign in with corporate credentials and access is removed when they leave. Confirm how roles map from directory groups to buyer permissions, and how the platform handles session timeouts and multi-factor policy. Account data is personal data, so document where it is stored and who can export it.
Analytics and event data
Analytics needs clean order, session and product events in one place, while privacy rules limit what can be sent to third-party tags. Decide whether events flow through a server-side pipeline, a tag manager, or your own data warehouse, and who approves each tracking tag. Confirm how the platform emits order and refund events, whether consent choices suppress tracking, and how long raw logs are retained. Reconcile analytics revenue against the ledger monthly so disagreements are found early.
What does an open source ecommerce implementation cost and require?
What drives the total cost of an open source ecommerce platform?
The software license is usually the smallest line item, because Magento Open Source (OSL-3.0), WooCommerce (GPL version 2 or later), Saleor (BSD-3-Clause), Medusa (MIT, excluding Enterprise Edition materials), and Sylius (MIT) can be downloaded without a license fee. The larger costs are the people and services around the platform: discovery and architecture, theme or storefront build, payment and ERP integration, hosting and staging environments, security testing, and ongoing upgrades. Ask every bidder to price these as separate lines and to state which costs recur each year. Enterprise add-ons, extension subscriptions and third-party connectors are also recurring and should appear in the quote.
How long does an open source ecommerce implementation usually take?
Duration depends on catalog size, the number of integrations, whether the storefront is headless, and how much custom checkout logic is needed. No duration can be stated for this topic without a written scope, so treat any timeline offered before discovery as a guess. A fixed-scope engagement should list discovery, build, integration testing, performance testing, security testing, migration rehearsals and launch as separate phases with agreed exit criteria. Ask for the critical path and which phase depends on the ERP or payment provider's sandbox access.
What does the client team need to provide during the project?
The client team must supply decisions and access that no implementation partner can make alone. This includes a product owner who signs off on catalog rules, tax and shipping logic, and returns policy; ERP and warehouse specialists who confirm data mappings and order flows; and an identity or security contact who approves admin access, single sign-on and network paths. Payment provider accounts, sandbox credentials, product data exports and a named approver for each go-live gate are also needed early. Projects slow down most when these owners are not named before the build starts.
Can our own team run an open source ecommerce platform after launch?
Yes, if the team has skills in the chosen stack and accepts the operating load. Running the platform means applying security patches and version upgrades, maintaining extensions, monitoring performance, managing backups and restores, and handling incident response for checkout. Magento Open Source and WooCommerce depend heavily on extensions and plugin compatibility, while headless options such as Saleor and Medusa shift effort toward API, storefront and service maintenance. Ask whether the team can perform a full staging upgrade within its own change window, and whether an external retainer for upgrades, support and continuous testing is needed.
How does PCI scope change with the ecommerce architecture?
PCI scope depends on how cardholder data flows through the system, not on the platform's license. A hosted payment page or iframe from the payment provider generally keeps the commerce application out of direct card handling, while custom checkout forms that touch card data widen the assessed environment. Headless storefronts add more components, including the API, storefront host, content service and CDN, and each one must be inside or excluded from scope on purpose. Ask your qualified security assessor to review the data flow diagram before the architecture is fixed. DeployOpen prepares systems and evidence for assessments but does not issue PCI DSS validation.
What makes migrating from Shopify or Salesforce Commerce risky?
The main risks are data loss, broken URLs and lost search rankings, mismatched order and customer histories, and integrations that behave differently on the new platform. Plan redirects from every indexed legacy URL, migrate customer and order history with clear rules about what is kept and what is archived, and run at least one full rehearsal before cutover. Ask the partner how they will reconcile inventory, pricing and tax between old and new systems during a parallel run. Migration effort rises sharply with custom Salesforce Commerce code or Shopify apps that have no open-source equivalent.
How much effort do upgrades take, and what should we budget for them?
Upgrade effort is recurring and depends on how many customizations, extensions and integrations sit on top of the core. Magento Open Source's latest stable tag is 2.4.9 in the magento2 repository, and WooCommerce's latest published version is 11.2.1, so both projects release updates that must be tested against your code and plugins. Ask for an upgrade runbook, a regression test suite that covers checkout and ERP sync, and a staging environment that mirrors production. Retainers for upgrades and continuous testing are available from DeployOpen, and their scope should be written into the same quote as the build.
Questions to ask an implementation partner before signing
Start with named experience, not general claims. Ask which of Magento Open Source, WooCommerce, Saleor, Medusa or Sylius the partner has deployed in production for a business of your type, how many of those deployments are still running, and who on the proposed team did the hands-on work. Ask for two references you can call and confirm the team members' roles in writing. Ask how they handle the platform's release line: Magento Open Source is licensed under OSL-3.0 and 2.4.9 is the latest stable tag in its repository, so ask which version they would deploy, how they track future tags, and how long they commit to patching it.
Next, ask how they scope PCI responsibilities. The partner should show, in writing, which systems fall inside your cardholder data environment, who operates the payment integration, and which controls belong to you, to the partner and to your payment provider. Ask whether checkout is hosted by the provider or runs on your servers, and what changes to the design that choice forces. For ERP integration, ask which system of record owns orders, inventory, prices and tax, how the integration is built (queues, scheduled jobs or direct API calls), and how failed messages are replayed and reconciled. For migration, ask for a written plan that covers URL redirects, customer and order history, and a cutover and rollback procedure.
Finally, settle the commercial and support terms. Ask which parts are fixed-scope with written deliverables and which are a retainer, and what each one excludes. Request a handover package that names the documents you will receive, such as runbooks, architecture diagrams, configuration records, test evidence and an access register. Ask for the support window, the response targets for security and production issues, the patch policy for both platform and extensions, and who pays for upgrades. Ask whether any part of the quote depends on a specific payment provider or hosting option. Do not sign until the scope, deliverables and timelines are agreed in writing and attached to the contract.
How DeployOpen scopes, builds, hands over and supports ecommerce platforms
DeployOpen plans, implements, customizes, integrates, hardens, backs up, upgrades, documents and supports self-hosted open-source platforms inside your own cloud accounts, data centers or air-gapped networks. For ecommerce, that means starting from your catalog, checkout, payment and ERP requirements, then delivering the platform under a written scope with evidence you can hand to your auditors.
How we scope
- Map catalog, checkout, tax, payment and ERP flows before choosing between monolith and headless architecture.
- Define PCI DSS scope in writing, including which hosts, payment redirects and integrations touch cardholder data.
- Agree deliverables, timelines and exclusions for the chosen platform in a fixed-scope document before work starts.
How we build
- Deploy the chosen platform, such as Magento Open Source, WooCommerce, Saleor, Medusa or Sylius, in your own environment.
- Build and test ERP, payment and shipping integrations, and run security testing on storefront, checkout and APIs.
- Migrate product, customer and order data from Shopify or Salesforce Commerce with reconciliation checks at each stage.
How we hand over
- Document architecture, configuration, integration maps, runbooks and the reasons behind key design decisions.
- Provide compliance evidence such as access records, change logs and backup test results for your assessors to review.
- Train your operations and development staff to run the platform, apply patches and manage third-party modules.
How we support
- Offer ongoing support, upgrades and continuous security testing for the ecommerce platform as a retainer.
- Track vendor release and security-fix windows so patching plans follow the supported versions of your chosen platform.
- Confirm backup, restore and upgrade procedures on a schedule agreed in the retainer scope.

