Choosing your ecommerce platform: the three real constraints
The choice of an ecommerce platform is decided on three constraints rarely discussed at the demo stage: the real complexity of your catalogue, your integration requirements with your systems, and the total cost over three years. Visible features rarely separate two serious platforms.
- The cost of a managed platform is paid in subscription and commissions; the cost of an open platform is paid in maintenance and skill.
- A catalogue with multiple variants or complex pricing rules eliminates more platforms than any other criterion.
- Migrating almost always costs more than expected, and the underestimation is about the data, not the design.
- Dropshipping is not a business model, it is a supply method, with its consequences for margin and service.
On this page
Managed ecommerce platform
A managed ecommerce platform is a service where hosting, security, updates and availability are handled by the provider, in exchange for a subscription and often a commission on sales. It stands in contrast to a self-hosted platform, where the company controls the code and the data but takes on maintenance, security and performance.
The three constraints that decide
Demos compare features: themes, apps, promotion tools. Every established platform has them, and that is never what blocks a project six months later.
Catalogue complexity
Number of variants per product, configurable products, pricing by customer or by quantity, unusual units of measure, multi-warehouse stock. It is the criterion that eliminates the most platforms, and you verify it by trying to create your most complicated product, not your simplest.
Integrations with your systems
Accounting, inventory, shipping, existing management system. A native integration is a project of days; an integration to build is a project of months, with a permanent maintenance cost. The question to ask is 'does it already exist', not 'is it possible'.
Total cost over three years
Subscription, commissions on sales, monthly third-party apps, theme, development, maintenance. A platform that is cheaper on subscription can cost more once commissions and apps are added up.
Third-party apps are the most underestimated line item on managed platforms. Eight apps at $30 a month add up to $8,640 over three years, often for functions that other platforms include as standard.
What to settle before the first demo
A platform is compared over thirty-six months, not on its monthly subscription. The line items that make the difference are invisible at the demo: commissions on sales, mandatory paid extensions, the cost of integrations with your systems, and the price of a migration if you change your mind. Three constraints decide, and none of them is a feature.
- How many references, variants and pricing rules does our catalogue have?
- Which internal systems must the platform exchange data with, and who pays for those connections?
- What is the total cost over three years, commissions and extensions included?
- What would a migration to another platform cost in two years?
- Can our customer and order data be exported in full, and in what format?
A solid answer puts a number on a total cost over three years and talks about your real catalogue. A hollow answer compares features, shows an interface, or stresses how fast the site can go live.
Choosing a platform commits a company for five years and is expensive to undo. A 90-minute consultation lets you validate the choice before signing, with a written summary you can hold up against the proposals you receive.
Comparing on what counts
| Criterion | Managed platform | Self-hosted platform |
|---|---|---|
| Initial go-live | Fast | Longer |
| Entry cost | Low | Medium to high |
| Recurring cost | Subscription and commissions | Hosting and maintenance |
| Very complex catalogue | Limited by the imposed structure | Flexible, at the cost of development |
| Data control | Partial | Full |
| Security and updates | Handled for you | On you |
| Changing providers | Full migration required | Possible without changing platform |
| In-house skill required | Low | Real and ongoing |
There is no good general answer. A company with no in-house technical resource and a simple catalogue has every reason to go managed. A company with a complex catalogue, unusual pricing rules or a high order volume often gains from controlling its platform, provided it takes on the maintenance.
The three-year calculation
The calculation structure below is filled in with your own numbers. That is what makes two proposals comparable.
| Line item | What to check |
|---|---|
| Monthly subscription | The tier actually needed for your volume |
| Commission on sales | The rate applied, and whether it drops with the native gateway |
| Transaction fees | Distinct from the commission, often forgotten |
| Third-party apps | A named list, with the monthly cost of each |
| Theme and customization | Initial purchase and planned adjustments |
| Integration development | Whatever does not exist natively |
| Maintenance and fixes | None when managed, real when self-hosted |
| Hosting | Included when managed, separate otherwise |
The commission on sales changes everything at high volume. One point of commission on two million in revenue is twenty thousand dollars a year, more than the full cost of a well-run self-hosted infrastructure.
What stays in-house is the honest description of the catalogue and the processes: the real number of variants, pricing rules by customer, management of multiple stock locations, the edge cases everyone forgets to mention before the migration. That is where budget overruns are created, and no provider can document it for you. What gets delegated is the rest: evaluating the platforms, costing over three years, the plan to carry over addresses, and executing the migration. A project that goes off the rails almost always started from an incomplete inventory of constraints.
Migrating without breaking everything
Migration is almost always underestimated, and the mistake rarely concerns the design. It concerns the data and the addresses.
| Element | Risk | Precaution |
|---|---|---|
| Product and category addresses | Loss of positions and links | Mapping table and complete 301 redirects |
| Order history | Loss of customer lifetime value | Migration or usable archiving |
| Customer accounts and passwords | Forced re-registration, loss of customers | Account migration, reset communicated |
| Product reviews | Loss of accumulated social proof | Export and re-import before the switch |
| Variants and attributes | Incompatible structure between platforms | Mapping done before, not during |
The switch is done with the redirects ready and applied the same day, never after. Every day of gap is a day when search engines and visitors hit errors on addresses that used to rank.
The dropshipping case
It is not a business model, it is a supply method where the supplier ships directly to the customer. It removes stock and tied-up capital, and it transfers to the merchant three problems it does not control.
Shipping times, the real quality of the products and returns handling depend entirely on a third party, while the commercial responsibility and the reputation stay entirely with you. In highly competitive categories, the margin is further compressed by sellers offering exactly the same catalogue.
It remains defensible for testing demand before investing in stock, or for rounding out an existing range. It is far less so as the foundation of a business, because it builds no asset of your own.
A platform is not chosen for what it can do. It is chosen for what it will prevent you from doing in three years.
Falia analysis gridBefore you decide
The address architecture stakes are detailed in ecommerce SEO. Profitability after the sale is covered in ecommerce conversion and retention. Execution is covered on the eCommerce page.
This point sits inside the plan described in the marketing plan and its budget.
Choosing a platform is a lever of the Improve your site's conversion goal.
Already running a marketing team? See how we plug in as reinforcement on online commerce.
Frequently asked questions about ecommerce platforms
Which platform should you choose to sell online?
It depends on three constraints: the complexity of your catalogue, the integrations needed with your existing systems, and the total cost over three years. The features visible in a demo rarely separate two established platforms.
Managed or self-hosted?
Managed suits companies with no in-house technical resource and a simple catalogue. Self-hosted becomes advantageous for complex catalogues, unusual pricing rules or high volumes, provided you take on the maintenance.
What is the real cost of a platform?
You have to add up the subscription, the commission on sales, the transaction fees, the third-party apps, the theme, integration development and maintenance, over thirty-six months. Third-party apps are the most underestimated line item.
What do you risk during a migration?
The loss of positions and links if the addresses are not redirected, the loss of the order history used to calculate customer value, and the loss of accumulated product reviews. The mistake almost always concerns the data, not the design.
Is dropshipping profitable?
It is a supply method, not a business model. It removes stock but transfers to the merchant shipping times, quality and returns it does not control. It remains useful for testing demand or rounding out a range.
When does the commission on sales become an issue?
At high volume. One point of commission on two million in revenue is twenty thousand dollars a year, more than the full cost of a properly maintained self-hosted infrastructure.
- Google Search Central, Migrate an ecommerce site to a new website, accessed July 2026.
- Google Search Central, Core Web Vitals, accessed July 2026.
- Thirty-six-month calculation structure: Falia working framework, revised in July 2026.
Gabriel almost always takes your first call and carries out your audit. He builds the strategy starting from your growth goal: where to put your budget, which market to test and how to connect each lead to a real sale in your CRM. He mainly leads engagements for three goals: Optimize the profitability of your digital campaigns, Develop a new market, and Generate demand and growth. With Geneviève, he also works on organic search (SEO), AI visibility (GEO) and conversion rate optimization (CRO). The sales a Google Ads or Meta Ads campaign brings in depend on the page that receives the click. He writes mainly about marketing strategy, paid advertising and measurement.
About Falia →
SEO
Paid advertising
Conversion
AI visibility
Web design