The Rise of the Specialist in an Era of Industry Consolidation
Why expertise is increasingly separating from infrastructure ownership
The enterprise technology industry continues to see consolidation. Teleco providers are combining operations. Security vendors are expanding into adjacent categories. Technology platforms broadening their portfolios. Systems integrators continuing to build scale and acquire new innovators.
On the surface, the direction appears obvious: bigger platforms, broader capabilities, and fewer strategic suppliers.
But another trend is happening at exactly the same time.
Specialists are becoming more important
That might seem contradictory. I don’t think it is.
In my first article in this series, the next era of enterprise networking had already begun, I argued that competitive advantage in enterprise networking has moved away from infrastructure ownership towards operating model, software, automation and outcomes.
In the second, history doesn’t repeat, but telco mega-mergers often rhyme, I explored why execution determines whether that strategic potential actually becomes client value or struggles to deliver real world benefits as large mergers languish.
Most recently, fully managed or DIY? Stop and consider the third way, I made the case that enterprises no longer need to choose between fully managed and DIY in-house. Co-management allows responsibilities to be allocated to whichever party is best equipped to perform them.
Those shifts lead naturally to another question:
If enterprises can increasingly separate the infrastructure they consume from the expertise required to operate it, where should that expertise sit?
I think the answer is changing.
Why specialist network expertise used to live inside the telco
For much of the history of global enterprise networking, the large telecommunications provider was the natural home for specialist network expertise.
That made sense. The telco owned much of the infrastructure. It operated the backbone. It employed the engineers. It had relationships with local carriers. It maintained the operational platforms and provided the managed service wrapped around everything.
For a multinational enterprise wanting global connectivity, buying the infrastructure and the expertise together was often the simplest model.
The two were intrinsically linked. If you wanted a sophisticated global network, you went to the organisations that owned sophisticated global networks.
But as I argued in the first article in this series, that relationship has been steadily breaking apart. Physical network ownership is no longer the primary source of differentiation. Software, security, automation, engineering and operating model have moved much higher up the value chain.
The expertise hasn’t disappeared. It has become increasingly independent of the infrastructure.
How software-defined networks separated intelligence from transport
SD-WAN was an important turning point. SASE then accelerated that fundamental change.
Previously, the connectivity service and much of the intelligence controlling it were tightly connected. SD-WAN helped separate those layers. The underlay could increasingly be sourced independently from the software-defined management layer sitting above it – the “overlay”.
The same pattern has continued.
Security became increasingly and inevitably cloud-delivered. Observability platforms created visibility across multiple providers and technologies. APIs allowed infrastructure to be orchestrated rather than manually configured. Automation began replacing repetitive operational activities.
The result is an important structural change: Enterprises no longer necessarily need to buy expertise from the organisation that owns the underlying infrastructure.
In many cases, they don’t even need one organisation to provide all the infrastructure.
That changes the economics of the industry, but it also changes where specialist capability can thrive.
Infrastructure is becoming more interchangeable. Expertise is becoming more specialised.
There is an interesting paradox here. Many elements of enterprise infrastructure are becoming easier to source from multiple providers.
Internet connectivity is available from thousands of carriers/ISPs. Cloud infrastructure can be consumed globally. Security and networking platforms are increasingly software-defined. APIs allow previously isolated technologies to work together.
Yet the expertise required to turn all of those components into a coherent, resilient and secure global environment is becoming more specialised. A modern enterprise network might involve multiple connectivity providers, SD-WAN, cloud security, digital experience monitoring, cloud platforms, automation and increasingly AI.
Someone still needs to make all of that work together in harmony.
As I wrote previously, delivering a high-performing global network increasingly means coordinating a diverse ecosystem of connectivity providers and digital platforms into a single operating model.
The interesting question therefore becomes: Where is the best environment for that specialist expertise to develop?
I believe it is inside organisations whose purpose and mission is developing and applying that expertise. Ideally, doing so with passion.
Consolidation and specialisation can happen simultaneously
Large-scale consolidation is not inherently negative for enterprise customers. Scale creates real advantages. Large providers can offer substantial reach, procurement leverage, obvious financial strength, broad portfolios and large teams of operational resources.
For many customer requirements those capabilities remain relevant. A large national retailer or a regional government body for example.
But scale and specialisation optimise for different things.
Large organisations naturally need standardisation. They have broad customer bases, large technology portfolios, established operating processes and significant legacy environments to support.
Specialist organisations can optimise differently. They can concentrate engineering talent around fewer technologies. They can change processes more quickly. They can adopt new platforms faster. They can build automation around a narrower set of problems. They can work more closely with individual clients because those relationships represent a greater proportion of their business.
Neither model is universally better. The question is which characteristics matter most for the outcome being sought.
Scale matters where scale creates value. Specialisation matters where expertise creates value.
The mistake is assuming they are the same thing.
Enterprise IT has seen this movie before
This isn’t unique to networking. Look at the ecosystems that have developed around enterprise software platforms. An enterprise implementing ServiceNow doesn’t necessarily assume that the organisation supplying its cloud infrastructure should also be its best ServiceNow implementation partner.
The same applies across Salesforce, SAP, Microsoft and other major technology ecosystems.
Specialist consulting and implementation partners have developed alongside the large global systems integrators because customers often value deep platform expertise, implementation experience and agility.
The platform vendor builds the technology. Infrastructure providers provide infrastructure. Specialists help enterprises turn the technology into an operational outcome.
Enterprise networking increasingly resembles this model. The connectivity provider doesn’t necessarily need to be the SD-WAN specialist. The SD-WAN vendor doesn’t operate the global network.
The security platform doesn’t necessarily need to provide every operational service surrounding it. A SASE platform may still integrate with another specialist security function.
And the enterprise doesn’t need to develop every specialist capability internally. In fact, with the speed of modern business and the race to deploy AI to your core business proposition, it is often unwise to have too many non-core initiatives.
Instead, these capabilities can be assembled around an operating model.
Co-management makes this possible
This connects directly to the previous article in this series. Traditional outsourcing encouraged enterprises to look for one provider capable of taking responsibility for almost everything. That inevitably favoured scale.
If the requirement was global connectivity, hardware, security, monitoring, operations, carrier management and support bundled into one contract, only relatively large organisations could credibly respond.
Modern technology has changed that constraint.
Shared platforms, granular permissions, APIs, automation and common visibility allow enterprises and providers to work inside the same environment. Co-management means the boundary between customer and provider can be deliberately designed rather than inherited.
That also means the boundary between different providers can be redesigned. An enterprise can retain architecture and policy internally, use specialist partners for engineering and operations, select technology platforms independently and source connectivity from multiple providers.
The objective isn’t to create a complicated collection of suppliers. Quite the opposite.
The objective is to put each responsibility where it can be performed best, while creating an operating model that makes the whole environment function as one service.
That is fundamentally different from traditional outsourcing.
Why best-of-breed sourcing fails without orchestration
There is an obvious risk. Specialisation can create fragmentation. Five excellent specialist providers don’t automatically create one excellent service. An enterprise that replaces one cumbersome incumbent with a dozen disconnected suppliers may gain technical capability while losing operational simplicity.
Ergo the rise of specialists makes orchestration more important, not less.
Someone needs responsibility for how the ecosystem works together. There needs to be common visibility, clear accountability, defined escalation, integrated processes and agreement over who makes which decisions.
This is why I keep returning to operating model throughout this series. Technology alone doesn’t solve the problem. Neither does supplier selection.
The value comes from how the capabilities are assembled and operated.
Seven questions CIOs should ask when sourcing a network partner
Perhaps enterprise sourcing teams should therefore ask some different questions.
Instead of simply asking, “Which provider can supply everything?”, consider:
- Which capabilities genuinely benefit from provider scale?
- Where does specialist expertise create greater value?
- Which components should remain independent to preserve flexibility and negotiating leverage?
- Who owns the architecture and ensures the different components work together?
- Can specialist providers collaborate effectively with our internal team and other strategic partners?
- How quickly can each provider adopt new technologies and operating practices?
- Are we selecting providers because their capabilities fit the requirement, or because their existing contract makes procurement easier?
These questions lead somewhere quite different from the traditional search for a single global provider.
They start with the capabilities required and work backwards towards the ecosystem needed to deliver them.
A Reflection in Closing
Industry consolidation will continue. There are sound economic reasons for it, and large providers will continue to play a role in global enterprise technology. But consolidation shouldn’t be confused with concentration of expertise.
In fact, the opposite is true.
As infrastructure becomes more software-defined, modular and interchangeable, specialist expertise can separate from infrastructure ownership and flourish in organisations specifically designed around those capabilities.
That creates more choice for multinational enterprises. But it also creates a responsibility to think differently about sourcing. The objective shouldn’t be to select the biggest provider, nor automatically the smallest or most specialised.
It should be to determine where scale creates value, where specialisation creates value, and how those capabilities can be orchestrated into a coherent operating model.
The specialist skills once had to live inside the infrastructure provider. Now, they don’t.
And that may prove to be one of the most important consequences of the next era of enterprise networking.
