The Collapse of the Entrepreneurial Model

The success of Tier 1 service providers is in jeopardy due to the deepening innovation crisis in management and operations technology.

Every Tier 1 service provider today relies on the same dual technical and market strategy: 

1. Attain lowest cost of service by migrating to a multiservice packet-based network; and,
2. Ensure highest customer loyalty by bundling valuable services delivered through high speed access networks. 

Yet every Tier 1 service provider’s success is in jeopardy due to the deepening innovation crisis in management and operations technology. Next generation networks – whether ‘next generation’ means IMS, LTE or metro Ethernet – can only be deployed in scale if they are supported by a new generation of supporting operations support technology. That technology is, with few exceptions, not being developed today due to the collapse of the entrepreneurial model for innovation in network technology.

The Innovation Crisis
New network technologies make their way into the marketplace through a web of innovative young companies, their investors and the larger established players that sometimes resell their products (and, one might add, often acquire and eventually squeeze the life out of them). Just like a planetary ecosystem, there is an interlaced and interlocking set of relationships here that ensure the evolution, growth and prosperity of the technology species.  Without the good health of this ecosystem, there would be no Cisco, no Ethernet, no Apple or iPhone, no Google and of course no modern network management or operations support software. 

Today this ecosystem is deeply sick. The mechanism that has for decades acted as a ‘wetlands’ for early stage network technology companies has gone toxic. The crisis predates the current recession; it began in the technology implosion of 2000-2001. Venture investment in virtually every sector had, of course, reached levels that could only be justified by willful speculative blindness to the realities of economics, valuation and normal market function. 

The ensuing and predictable collapse drove venture investment as a whole back to levels that were historically rational. By 2004, the market had continued a normal growth pattern, as if the bubble and burst had never occurred. But investment in network technology never recovered, and has continued its decade-long decline into the present. In fact, 2008 network technology investment has reached a low that only barely exceeds the level spent in 1996. The little investment that has continued serves almost exclusively to support later stage investments in companies that were started many years earlier. Enterprise creation in the network technology sector is, for all practical purposes, a dead zone. 

Data Source:  Moneytree Report, PWC & NVCA

Next Generation Networks:  New Management Requirements
To understand why this must be a concern for every Tier 1 network service provider, consider just one set of requirements for operations in a next generation network context.

A single service, such as consumer voice, once relied on a monolithic network. One class of network device (stored program control switch) in a small number of vendor and model variations (Lucent #5ESS and Nortel DMS-100, say) were the primary programmable engines that delivered service and required operator intervention, much of it delivered (locally or remotely) through a dedicated console. A single device delivered service to a population of some 10,000-20,000 subscribers, on average.

In one current Tier 1 IMS network, consumer VoIP is delivered through a combination of 26 discrete network devices, each of which may be replicated physically many times in order to provide service within a given market area. Each device requires a specific operating software version and patch level; each may have from several hundred to several thousand software-driven parameters that may or may not need to be set with reference to a predefined operations and procedures model (which is itself in constant evolution). Each device has its own logon procedure, password and security requirements. The ratio of devices to subscribers and services delivered is two to three orders of magnitude higher than in the POTS generation. 

The problem of how to commission new devices in the network; how to distribute, backup and restore new versions of software; how to audit software configuration, compare to a reference model and flag discrepancies that may be service-affecting; and how to manage the security requirements for access across the thousands of devices that deliver service in a given market area: there is no solution to this operational problem in the element management software delivered by the device vendors; there is no solution to this operational problem in the established network management or provisioning solutions that evolved over the last decade in response to the rapid growth of DSL or digital wireless. 

This is not an IMS problem. The same pattern applies in the evolution from 3G to 4G LTE wireless. Replace relatively dumb 3G base stations with intelligent IPv6 routing solutions, and you have the same multiple-order-of-magnitude increase in management load. It’s not just an IMS and LTE problem. Replace relatively simple TDM optical transport systems with more complex Ethernet solutions, try to manage them to the same level of predictable metrics, and you have yet a different permutation on new operational management requirements. The process of network disaggregation has fundamentally changed the requirements for management solutions.

Nakina Systems, because it is a rare exception to the overall business cycle (a post-crash OSS innovator with a Tier 1 footprint), has developed the capability to address this “domain control” management problem, but this is the exception that proves the rule. There are dozens of unfulfilled operations requirements that stand between the promise of a converged, efficient, multiservice network and very little occurring in the way of enterprise creation or even product innovation to address them.  

Great Companies That Will Never Get Started
There is no shortage of big problems to solve in network operations and telecom applications more generally. A great little company could be built around the needs for next generation transport network planning and design, incorporating wireless backhaul regrooming, TDM to Ethernet transition and LTE network design. Another one could rewrite the rules in the inventory market by solving the ‘data currency’ problem and replacing a static network inventory with a virtual one – combining stored and in-network data transparently. Still another one could do for applications content and settlements in an open market what the Apple App Store does in a closed garden. But until the entrepreneurial model returns to some kind of balance, these needs aren’t going to form the seeds of new enterprises.