Build vs Buy Is the Wrong Question. Decide What You Actually Need to Build.

Buying or building software is rarely an all-or-nothing choice. How to evaluate capabilities, integration and operational responsibility, illustrated by Schubert's recruiting funnels.

Key points

  • Evaluate build versus buy at the capability level rather than choosing one approach for an entire platform.
  • Compare subscriptions and development on equal terms, including integrations, operations and switching costs.
  • Schubert's recruiting funnels connect Umantis, Sanity and Remix rather than recreating applicant tracking.
  • Hybrid architectures need clear data ownership, tested failure handling and realistic exit options.

Buy software or develop it in-house? For many software initiatives, that is the wrong level of decision. Established capabilities need not be rebuilt. Where business rules, data flows or the ability to change a workflow are distinctive, a tailored component may be justified. The practical build-versus-buy decision is often made capability by capability, not for an entire application.

Start with the workflow. Which capabilities are already covered? Where does the business need control over changes? And who will own the interfaces and ongoing operations?

A license fee is not a cost comparison

Make-or-buy discussions often compare a subscription fee with an estimate for custom development. The comparison is incomplete. Purchased software may require implementation, configuration, integration, external support and future migration. Custom software carries testing, security maintenance, monitoring, documentation, hosting and further development costs as well as its initial build cost.

Compare like with like: the same functional scope, evaluation period, expected change rate and ownership model. Microsoft discusses control, customization, time to market, available expertise, support and operating costs when weighing development against existing solutions. AWS recommends evaluating total cost of ownership at component level, including operational and management costs. Both are cloud-architecture sources; the underlying cost questions also apply when choosing business software.² ³

Schubert: develop the missing capability, not another platform

In our Schubert Group recruiting-funnel project, the applicant tracking system and website were already in place. Job listings were managed in Umantis, while recruiting campaigns relied on a separate funnel platform with agency support. This added tool-management work, coordination and an additional workflow.¹

Job data is imported from Umantis into Sanity, where the HR team can assign positions to recruiting funnels and manage campaign content. Remix delivers those funnels through the existing website using its components and campaign tracking.¹

We developed the necessary integrations and funnel functionality within the existing architecture. Building a second applicant tracking system would not have solved the actual problem.

Control matters when requirements change

Architecture choices are about responsibility as much as technology. Who owns the business rules? Which system is authoritative for each data set? Who can change an interface and keep the integration running? These boundaries should be explicit even when most of the functionality comes from standard products.

The real test comes when something changes: Can the data be exported? Can integration behavior be tested? Can teams adapt business rules without commissioning every adjustment externally? Owning code provides limited freedom if a critical vendor API remains undocumented or proprietary.

Four options, not two: Buy, Integrate, Build, Replace

Buy

Choose an existing product when the workflow is standardized, the functional fit is good and the vendor can reliably handle operation, security and product development. Applicant tracking, invoicing and collaboration are common candidates.

Integrate

Connect working systems when what is missing is the flow of information or a specific process step. This requires usable APIs, clear data ownership, error handling and defined operational responsibilities.

Build

Custom development makes sense when a capability differentiates the business, cannot be configured sensibly or needs a level of control and adaptability that standard products do not provide. Maintenance and operations must still be budgeted. For scope, integration and cost considerations, see our custom software development service.

Replace

Replace an application when its limitations block necessary integrations or changes and continued workarounds cost more than a controlled transition to a better-supported alternative.

Build and buy moves risks; it does not eliminate them

A hybrid system creates more boundaries. Data models and APIs change, permissions must work across products, and failures can be difficult to attribute. Without someone accountable for monitoring and modifications, integration becomes a risk in its own right.

Define the failure paths before deployment. What happens when an import fails? How will records be retried? Who detects and resolves a broken synchronization? The ability to switch providers later should also be considered while interfaces and data contracts are being designed.

How we prepare a build-versus-buy decision

We describe the actual workflow first: triggering events, data sources, user roles, security requirements and expected changes. That makes the critical capabilities and architecture assumptions visible before specific products are shortlisted.

Where assumptions are uncertain, we test them with a small API prototype, a representative data import or a thin, end-to-end slice of functionality. We document the constraints, ownership and realistic exit paths alongside the findings.

The decision is not permanent. A changed licensing model, API limitation or security requirement can invalidate the original assumptions and justify another evaluation.

Conclusion: build only what the business needs to own

The strongest build-versus-buy decision does not maximize custom code. It gives each capability a clear role, manageable change paths and accountable operations.

Through Consulting & Diagnostics, we evaluate the architectural options. With Platform Engineering, we implement the functionality and integration that genuinely warrant custom development.

Sources

¹ Valanor, “Schubert Group: Recruiting Funnels with Sanity and Remix”, project reference, 2026: https://valanor.co/en/projects/schubert-recruitment-funnels-sanity-remix

² Microsoft, “Architecture strategies for getting the best rates from providers”, Azure Well-Architected Framework, accessed 22 November 2026: https://learn.microsoft.com/en-us/azure/well-architected/cost-optimization/get-best-rates

³ Amazon Web Services, “COST05-BP03 Analyze the total cost of ownership of each component”, AWS Well-Architected Framework, accessed 22 November 2026: https://docs.aws.amazon.com/wellarchitected/latest/framework/cost05-bp03.html

Consulting & Diagnostics

Which capabilities should you buy, integrate or develop?

We assess requirements, interfaces, operating risks and cost drivers to establish a defensible build-versus-buy decision for your project.

Schedule an Initial Technical Consultation

Aleksandar Basara

Managing Director

aleks@valanor.co

+49 (0) 711 16 44 6 83

Lindenspürstrasse 22, 70176 Stuttgart

Aleksandar Basara