Enterprise software sourcing often begins with a simple question:

Which product best meets our requirements?

The logic is sound. Specialist products can provide deeper functionality, closer business fit and faster innovation than broader alternatives.

But product-by-product optimisation rests on a larger assumption: that the strongest application in each category will produce the strongest enterprise technology environment.

That assumption is incomplete.

The product decision and the portfolio decision cannot be separated. An application’s value depends not only on what it does individually, but also on how it affects workflows, data, cost, control and the organisation’s ability to change.

What Has Changed?

Signal 1: Portfolio Economics Extend Beyond Product Economics

Product evaluations typically compare functionality, implementation and price.

The enterprise bears a wider cost.

The US CIO Council’s application-rationalisation methodology evaluates business value, technical fit and total cost of ownership. It also examines dependencies, interoperability, duplication and security. (US CIO Council)

These costs are difficult to see. In 2024, the US Government Accountability Office found that none of nine selected agencies had fully determined whether their five most widely used software licences were over- or under-purchased. (GAO)

A functionally superior product may still be the right choice. Its full portfolio cost must nevertheless be visible.

Signal 2: Value Crosses Application Boundaries

Enterprise outcomes frequently depend on shared processes and data rather than one application.

MIT CISR associated mature technology platforms with lower IT unit costs and greater reliability. Later research found that architecture capabilities produced better performance only when paired with effective management practices. (MIT CISR)

A study of 1,902 US hospitals found that single-vendor electronic-health-record environments were more likely to support both clinical and organisational evaluation than best-of-breed environments. The result was observational and sector-specific, but it illustrates how portfolio design can affect system-wide capability. (Mayo Clinic Proceedings)

AI may strengthen this consideration. UK research found integration, scaling and data complexity among the barriers reported by AI users. The evidence remains emerging and does not demonstrate a single-vendor advantage. (DSIT)

Signal 3: Control Creates a Two-Sided Risk

Each material supplier relationship brings due diligence, contracting, monitoring and incident-planning responsibilities. NIST treats these as lifecycle responsibilities rather than selection-stage checks. (NIST)

Consolidation creates the opposite risk.

The EU’s Digital Operational Resilience Act requires financial institutions to assess provider concentration, substitutability and exit arrangements for critical services. (DORA)

More vendors do not automatically mean more resilience. Fewer vendors do not automatically mean more control.

The Decision That Needs Rethinking

The choice is not best-of-breed versus suite.

It is whether each application is evaluated only as a product—or also as part of the enterprise portfolio.

Functional capability remains essential. It should be considered alongside lifecycle economics, dependencies, data, security, concentration and reversibility.

This does not replace product-first with suite-first, or specialisation with standardisation.

It widens the decision.

What Hasn't Changed

Best-of-breed products can still provide capabilities that broader platforms cannot match. For strategically differentiating functions, that superiority may justify substantial integration and governance costs.

Open interfaces and portability requirements may also make heterogeneous portfolios easier to operate. The EU Data Act introduces switching and interoperability requirements, although the operational effects are not yet established. (EU Data Act)

A strong portfolio can contain multiple specialist products.

The requirement is not uniformity.

It is coherence.

Executive Questions

  • What enterprise value justifies a product’s superior functionality once its full portfolio effects are included?
  • Which dependencies, concentration risks and governance obligations will this decision introduce?
  • How easily could we change or exit the product without weakening critical workflows, data or controls?

Three Signals to Monitor

Lifecycle Economics

The realised cost of integrating, operating, changing and eventually replacing applications.

AI at Scale

Measured production outcomes across integrated and heterogeneous application environments.

Interoperability and Exit

Real switching costs, tested exit capability and changes in provider concentration.

Bottom Line

Best-of-breed is not the problem.

The problem is treating product selection as an isolated decision.

The better question is not simply:

Which product is best?

It is:

Will this product make the enterprise portfolio better?

References

  1. US CIO Council. View source
  2. US Government Accountability Office. View source
  3. MIT CISR. View source
  4. Mayo Clinic Proceedings. View source
  5. UK Department for Science, Innovation and Technology. View source
  6. NIST. View source
  7. Digital Operational Resilience Act. View source
  8. EU Data Act. View source