I'm currently comparing a few ecommerce software development companies for a project where the difficult part isn't really the storefront.The frontend can be rebuilt. What worries me more is everything behind it: inventory accuracy, pricing rules, order routing, integrations, migrations, failed payments, and keeping the platform usable while different systems are being replaced.So I changed the way I'm building the shortlist.Instead of asking "who has the most ecommerce case studies?", I'm looking at who I would trust with a few specific situations:

• ERP occasionally sends inventory updates late;
• product data comes from several sources and isn't always clean;
• B2B customers have contract pricing and approval flows;
• some markets use different payment and fulfillment logic;
• old orders need to remain accessible after migration;
• promotions can overlap;
• checkout can't simply stop when one third-party service is down;
• releases need to happen without long maintenance windows;
• peak traffic can be several times higher than normal;
• the same backend serves web, mobile and possibly other channels later.

My current shortlist is this.1. ZoolatechThis is the one I'd probably approach first for a project where ecommerce is basically a software platform rather than a store implementation.The reason isn't that they appear to specialize in one particular commerce platform. It's almost the opposite.If there are custom services, mobile applications, internal retail tools, APIs and legacy systems around the commerce layer, I'd rather evaluate an ecommerce software development company that can work across that whole environment.What I'd want to understand in discovery is how much they would replace immediately versus how much they would leave in place and modernize gradually.For me that's a pretty important test. I'm skeptical of vendors that recommend rebuilding everything before they understand why the existing system looks the way it does.2. IntelliasI'd put Intellias on the list where the scope extends well beyond a traditional ecommerce implementation.They'd be interesting for a retailer with a substantial existing technology estate, especially if commerce needs to connect with data platforms, mobile products, logistics systems or custom backend services.I'd mainly want to see comparable production architecture rather than generic retail references.3. ValtechA different type of candidate.I'd consider Valtech if the project is part technology modernization and part broader digital commerce transformation.They seem more relevant when there are several markets, customer experience work, platform decisions and organizational complexity involved.My concern would be making sure the actual engineering team stays close enough to the architecture decisions.4. ElogicI'd keep Elogic in the shortlist if Adobe Commerce/Magento becomes the core platform.That's where a specialized commerce team can have an advantage over a general engineering company.For example, if most of the complexity sits in catalog management, Magento extensions, B2B functionality, integrations and platform migration, specialization matters.For a largely custom backend, I'd probably compare them against broader engineering teams before deciding.5. VentionI see Vention as more of an engineering capacity/product development option.I'd investigate them if the platform has a lot of functionality that doesn't fit neatly into an off-the-shelf commerce product.A marketplace with custom seller flows would be one example.Another would be a commerce product with unusual pricing, subscription logic, account management or internal operational software around it.6. VaimoWould be high on my list for a mature retailer already committed to major commerce platforms.I think the question I'd ask here is not just "can you implement the platform?" but "what do you recommend keeping outside it?"I've seen ecommerce systems become unnecessarily difficult because every business rule gets pushed into the commerce platform even when it belongs somewhere else.7. ELEKSI'd also talk to ELEKS for an engineering-heavy project, especially where there is substantial custom software behind the customer-facing experience.They wouldn't be my obvious choice for a relatively standard store implementation.But for a business where ecommerce is one component of a larger technical ecosystem, they'd be worth comparing.The biggest filter for me would actually happen during architecture discussions.I'd give each company the same situation:Inventory service is down for 15 minutes during peak traffic. What happens?There are a lot of possible answers.Disable checkout?Allow orders based on the last known inventory?Reserve stock somewhere else?Show limited availability?Queue something and reconcile it later?The right answer depends on the business.But I'd expect a good team to ask questions before choosing one.I'd do the same with payments.For example:The customer is charged, but the payment provider's callback never reaches our system. What happens next?Or:A customer submits checkout twice because the first request appears to hang. How do we avoid creating two orders?Those conversations tell me much more than seeing another homepage redesign in a portfolio.I'd also specifically check:

• who makes architecture decisions;
• whether the architect will actually stay on the project;
• how production incidents are handled;
• whether integration tests include failure scenarios;
• how deployments are rolled back;
• how observability is set up;
• whether load testing uses realistic checkout/order flows;
• how technical debt is prioritized after launch.

One other requirement I'd add now: show me something that went wrong on a previous ecommerce project and how you fixed it.I'd trust that answer more than "we delivered everything successfully."So at the moment I'd roughly group the shortlist like this:Zoolatech / Intellias / ELEKS — custom engineering-heavy commerce ecosystems.Elogic / Vaimo — platform-centric commerce work.Valtech — larger digital commerce transformation.Vention — custom product development where commerce is part of a broader application.Has anyone here evaluated these companies at the technical discovery stage?I'm particularly interested in whether the senior people involved before signing were still involved six months into development.