The Rise of the PaaS Aggregator
And Why PAs will Leave SIs Behind
For nearly three decades, system integrators (SIs) have dominated the IT outsourcing market, thriving on a model built around labour arbitrage. These firms leveraged large, low-cost offshore workforces to design, build, and maintain enterprise systems, operating on a cost-plus approach where success was measured in billable hours rather than actual business outcomes. This model worked well in an era where enterprise software landscapes were fragmented, integration was complex, and custom development was the only viable way to bridge the gaps between different systems. Large organisations in particular, let’s say banks or governments, relied on SIs not only for their technical expertise but also for access to a scalable, specialised labour force capable of managing these bespoke integration challenges at scale.
But the fundamental assumptions underpinning this model no longer hold true. That era is ending.
Take Your Reading on the Go! 🚀
Did you know? The Substack app lets you listen to this post as an audio version. That’s perfect for your next 15-minute commute or lunchtime walk. Download the app and tune in! 🎧 Thanks for reading or listening!
The IT services industry is undergoing a seismic shift. It will not be incremental but foundational. It is not just that system integrators are being displaced. Rather, the function of systems integration is being redefined. Traditional SI-driven custom integration work is giving way to pre-integrated platform ecosystems, where technology service providers act as orchestrators and platform aggregators (PAs) rather than builders of bespoke solutions.
Unlike their predecessors, these firms will not need to rely (solely) on vast armies of developers and offshore coding factories. Instead, they will focus on assembling domain-specific workflows within powerful Platform-as-a-Service (PaaS) ecosystems. This will vastly reduce the need for large-scale custom development and shift the value proposition from labour-intensive software projects to AI-driven automation and preconfigured business processes.
The implications, especially in Asia Pacific, are considerable.
The collapse of the traditional SI model begins with the breakdown of its underlying economic logic. Labour arbitrage, the idea that a company's competitive advantage comes from shifting work to lower-cost regions, no longer holds the same power in a world where intelligent automation and configurable platforms can do much of the heavy lifting.
Yes, it is still cheaper for an Australian or Singaporean government or multinational to outsource to India or Vietnam, but the equation has changed. The unit cost savings are shrinking, and more importantly, the long-term liabilities are growing. Traditional SI-driven projects often function like long-term debt on an organisation's technology balance sheet. They provide short-term gains but accumulating structural burdens in the form of rigid architectures, high maintenance costs, and expensive future upgrades.
Just as financial analysts scrutinise the sustainability of long-term debt, organisations must now assess whether their technology investments are creating enduring value or merely deferring future costs. The appeal of SI-driven custom integration, once viewed as an asset, is now being reevaluated as a liability. And one that many organisations are eager to retire.
This transformation represents more than just a shift in tooling. It is a fundamental change in how organisations account for technology investment and risk. The previous assumption that large-scale development teams were necessary to build and maintain business systems has been eroded by the rise of low-code and no-code solutions, which allow organisations to create sophisticated enterprise applications without needing to write and maintain millions of lines of bespoke code.
This reallocation of technical debt is akin to moving from long-term liabilities to flexible, on-demand financing models where software adapts to business needs without locking organisations into costly, multi-year development cycles. AI-powered automation has accelerated this trend, removing the need for manual integration and maintenance work that once justified high-value outsourcing contracts.
Where once companies had no choice but to borrow against their future agility, by signing long-term SI contracts to fund the development and upkeep of rigid, custom-built systems, today, they can deploy modular, self-optimising platforms that reduce both cost and complexity. The result is a more liquid, adaptive technology portfolio, one that shifts from heavy debt financing (high-cost, long-term SI contracts) to pay-as-you-go efficiency (automated, configurable platforms).
For traditional SIs, this is an existential problem. Their business model was built on enterprises treating technology as a high-cost, capital-intensive investment. Our generation was defined by solutions that required deep technical expertise, complex integrations, and long-term support contracts. But as more businesses shift towards subscription-based, AI-augmented, and low-code solutions, the justification for these legacy contracts is collapsing. The SI model, once an asset on the enterprise balance sheet, is becoming a liability. And it is one that fewer organisations will be willing to carry.
Isn’t PaaS Just the New ERP?
For years, the core work of SIs has not been about building entire ERP systems but rather about designing and maintaining the business applications that operate between, around, and on top of ERPs and other enterprise platforms. Their role centered on constructing workflows that automated business processes and linked various enterprise applications, ensuring that disparate systems could function together in a cohesive manner.
They were responsible for developing system-to-system integrations, connecting ERP suites such as SAP, Oracle, and JD Edwards with other business-critical applications to enable seamless data flow. Additionally, they played a key role in bridging workflows across multiple systems, ensuring that complex business processes functioned end-to-end across fragmented IT environments.
Beyond the back-end logic, SIs also built and maintained the presentation layer, designing the user interfaces and experience frameworks that allowed customers and employees to interact with these workflows and business applications efficiently (see The Portal Wars Are Back).
Historically, these elements were heavily custom-built. SIs and their customers relied on a combination of bespoke middleware, custom logic, and lengthy DevOps pipelines to manage and update services. Their value came from building and maintaining these complex, tightly coupled integrations, which required constant updates and support.
Yet even when SIs were "integrating" solutions, they often did so through stack-based development, rather than creating true plug-and-play interoperability. This meant that, over time, many organisations accumulated an increasingly brittle patchwork of customised software deployments, requiring continuous SI intervention for upgrades, maintenance, and security patches.
This entire model is now being upended by PaaS.
The need for custom development to integrate systems and automate workflows is rapidly disappearing as organisations transition to platform-native solutions that reduce complexity and increase agility. Instead of writing bespoke code to link different enterprise applications, businesses can now configure and orchestrate very complex workflows directly within PaaS environments. These platforms provide built-in automation capabilities that replace traditional development-heavy approaches with configurable logic and low-code process modelling.
At the same time, system-to-system integrations no longer require custom middleware or extensive development effort. Leading PaaS ecosystems now offer prebuilt connectors, API-driven automation, and plug-and-play integrations, allowing applications to communicate seamlessly without the need for complex coding. The reliance on dedicated SI teams to build and maintain these connections is being significantly reduced. Modern platforms are now capable of providing what we always wanted with ERP but could never have: native interoperability across business applications.
This is a fundamental shift. It is not simply about replacing ERP systems with PaaS, nor is it about writing entirely new enterprise applications within PaaS environments. Rather, it is about recognising that most “business applications” in the SI sense, things like workflows, system integrations, and UI layers, can now be built and maintained directly within PaaS ecosystems without traditional custom development.
This shift breaks the traditional SI model in multiple ways. First, it removes the need for bespoke integrations, eliminating one of the primary sources of SI revenue. Second, it makes ongoing maintenance easier and cheaper, reducing the necessity of long-term support contracts. Third, it shifts the skill set from software development to platform configuration, further reducing demand for the large offshore development teams that have historically been the backbone of SI operations.
For customers, this means faster development cycles, shorter project timelines, and lower overall costs.
By eliminating custom development, solutions can be deployed more quickly, reducing the complexity and risk traditionally associated with large-scale implementations. The ability to configure rather than code also accelerates testing and deployment, making it easier to iterate and refine solutions without prolonged development phases. Additionally, the reduced reliance on large offshore teams translates to lower operational expenses, while the simplification of maintenance reduces the need for costly long-term SI contracts. Ultimately, customers benefit from a more agile, cost-effective approach that gives them greater control over their technology investments, enabling them to adapt to business needs with far less friction.
With this shift will come a fundamental restructuring of the IT services industry.
System integrators have long operated on a linear growth model, where scaling required expanding headcount to take on more projects, driving revenue through billable hours. However, as automation and platform configuration replace custom development, this model no longer scales.
More importantly, it is becoming increasingly unappealing as a long-term investment. Industry analysts are already questioning its viability, and financial analysts will soon follow. The SI model has been sustained by an 8–10% profit margin, but maintaining that in a growth environment is becoming increasingly dependent on cannibalising a labour arbitrage market that is visibly collapsing.
As automation, AI, and platform-based workflows replace the need for large offshore teams, the SI business model is beginning to look like a legacy approach, much like what happened to traditional BPOs when SaaS and automation eroded their relevance. As revenue growth slows and margins come under pressure, the SI model will struggle to maintain investor confidence, leaving its long-term prospects in serious doubt.
Changes to the Partner Model
Another major shift is in the nature of vendor relationships. Traditionally, SIs maintained broad partnerships across a wide range of independent software vendors (ISVs), often working with dozens of platforms and technologies. However, platform aggregators (PAs) will need to operate differently, prioritising deep, strategic relationships with a select group of core PaaS providers rather than spreading their expertise too thin.
Despite this shift, PaaS providers continue to treat their long-term SI partners the same way they always have, without recognising the need for a different approach. The largest SIs should, at a minimum, be as technically proficient as the best PaaS employees themselves, yet too often, they remain mono-dimensional in both their understanding of the platform and their go-to-market strategies. To succeed, the PaaS leaders in this space must rethink their partner programs, ensuring that the next generation of service providers is fully aligned with the evolving needs of the ecosystem.
Unlike the cursory accreditation programmes of the past, successful PAs will invest heavily in platform-specific expertise, ensuring their teams have deep technical skills and high-level certifications such as Certified Technical Architect (CTA) and Certified Master Architect (CMA). The ability to configure, optimise, and extend PaaS capabilities should now be the primary differentiator in the IT services market.
The rise of embedded AI within PaaS further accelerates this transformation. As AI-driven workflow automation, predictive analytics, and machine learning become standard features of modern platforms, the need for traditional manual development continues to erode.
This shift does more than just improve efficiency. It fundamentally reshapes the nature of IT service delivery. The PA firms that thrive in this new environment will be those that embrace AI-enhanced automation and focus on delivering business value through intelligent platform configuration rather than relying on labour-intensive development.
For the global SIs that have long equated dominance with sheer headcount, this is an existential challenge. In the old model, size mattered. More people meant more capability, more influence, and more revenue. But in an economy where automation replaces manual effort, size without intelligence becomes a liability. The largest firms can no longer be considered the biggest silverbacks in the jungle. They are simply the slowest-moving targets.
Those that fail to adapt will not just struggle. They will be actively penalised by an economic model that rewards agility, intelligence, and efficiency over brute scale.
Case Study: Coforge
Many SIs are beginning to contemplate this shift. Some are even demonstrating early success. Coforge provides an interesting case study.
Coforge1 itself may not have always had a clear answer to what has driven its success over the years. When asked, their response has often been simply, "customers love us." While that may be true, the more important question is why? And that’s something they haven’t been able to fully articulate, at least to me.
While I wouldn’t say they deliberately set out to adopt the platform aggregator model, their recent progress, whether by design or adaptation, suggests they are naturally evolving toward a delivery approach that better aligns with the changing technology landscape.
Their strong presence in industries such as banking, financial services, insurance, transportation, and logistics reflects this shift. These sectors have seen traditional system integration approaches struggle to keep pace with evolving client demands, yet Coforge has managed to carve out a highly competitive edge. Their success doesn’t fit neatly into conventional IT services models, but when viewed through the lens of platform aggregation, their trajectory starts to make far more sense.
They have developed strategic partnerships with key PaaS providers, including Salesforce, Pega, ServiceNow, Duck Creek, Appian, AWS, Microsoft, and Google, aligning their services with the major workflow and hyperscaler platforms that are driving enterprise transformation. This alignment enables them to deliver solutions that integrate more seamlessly into modern enterprise ecosystems.
Beyond their technical strategy, Coforge’s real strength lies in its client-centric approach and focus on long-term value. They have built deep, enduring relationships with major clients such as British Airways, Virgin Australia, Arnott’s, and Coles. Not by competing on headcount or pricing but by delivering preconfigured, industry-specific solutions that drive faster, more scalable outcomes.
By leveraging prebuilt workflows within established PaaS ecosystems, they are minimising the need for extensive custom development while enhancing the efficiency and agility of their clients’ technology operations. This approach not only accelerates implementation but also reduces long-term maintenance burdens, making their services more sustainable and cost-effective in a rapidly evolving digital landscape.
And that, I believe, is the real differentiator. The challenge, however, is that Coforge invests little in marketing, which sometimes leads to a lack of clarity in its go-to-market messaging.
When messaging is developed entirely from within the company, without external validation, it can become insular and difficult to communicate effectively. Great technical and solution architects don’t always translate into great marketing strategists, and that gap is evident in how Coforge presents itself to the market. So at this point they are still positioning themselves as a challenger SI.
To really lock in on this model they have some work to do. And that has implications for them and their partners and their customers. In the past, a firm like Coforge may have used Salesforce for CRM, ServiceNow for ITSM, and Pega and Appian for general workflow automation. But today, each of these platforms are expanding their ambitions, positioning themselves as broader enterprise-wide solutions that challenge traditional ERP vendors like SAP.
Salesforce is no longer just a CRM. With Financial Services Cloud, Public Sector Cloud, and MuleSoft, it is making a serious play into ERP-like territory. ServiceNow, once seen purely as an ITSM tool, is aggressively expanding into customer service, HR, and enterprise-wide workflow automation into every corner of the enterprise.
For Coforge, simply being a PaaS partner is no longer enough. Service providers will have to make strategic decisions about which PaaS vendors they go deepest with, forming exclusive alliances that shape the new IT services landscape.
It is now incumbent on PaaS vendors to ensure that their platform becomes the foundation of choice for platform aggregators, offering incentives and building stronger partner ecosystems that make their solutions indispensable in this evolving market.
The bottom-line on all this: the future of IT services belongs to Platform Aggregators (PAs), not System Integrators (SIs).
This transformation is not a trend. It is a fundamental redefinition of how technology services are delivered. The days of labour-intensive, custom-coded solutions are numbered, replaced by a world where AI-driven automation, prebuilt workflows, and deep PaaS specialisation define success.
Organisations, and that includes end-user customer organisations, still operating under the SI model must start to adapt or risk irrelevance in a rapidly moving industry.
For full transparency, I recently attended the Coforge Analyst and Advisor Day as a guest in Sydney on March 11, 2025.