So, for existing CSPs the choice becomes about embracing innovation or running the risk of being displaced by new entrants. How do we know that displacement will occur?  Because it has been done before—see Western Union, Motorola, and others. The only way out is to embrace innovation.

Barrier to Overcome

Many CSPs have been talking about the importance of innovation for some time.  How come then, hasn’t it happened? Despite the best intentions, current ways of dealing with technology have created a barrier.

Innovation generally comes from small groups, either start-ups or small companies.  Current CSP approaches favor very large suppliers. A Silicon Valley venture capitalist tells the story of a CEO from a major telco complaining that he didn’t see enough start-ups. The VC says he told the CEO that they didn’t see many start-ups because they killed them.

The small number of very large vendors that provide CSP infrastructure can have senior staff groups work for free with CSP advanced technology groups, standards organizations, open source groups, and more while waiting for a very large comprehensive RFP (Request For Proposal). They have dedicated sales teams that “live” with their respective CSP. Then, they have specialized groups to respond to the large RFPs. These large vendors are very happy to do this because they see this process as a barrier to entry from new competitors. 

The large vendors are able to recover all the expense inherent in this system in their large RFP-based sales. In terms of competition between these large vendors, they all have the same expense profile, so they do not compete based on these expenses. In other words, for the large vendors, this is simply a cost of doing business.

CSP’s realize these costs in what they perceive as inflated margins on the products they buy through these large RFPs. They respond by trying to squeeze as much out of the large vendors as possible through unpaid activities. Unfortunately, the CSPs apply the same pressure for unpaid activities on the start-ups and small companies that do not have the requisite financial structures to survive and stay in this game. 

After one such high-priority, highly successful proof-of-concept, a start-up completing a successful PoC (that could not be accomplished by the large vendors) asked a senior CSP advanced technology manager for a paid lab test engagement. The manager responded, “Why should I pay you anything? I already have everything I need.”  Do you see the barrier to innovation?

Small companies do not have the financial structure to do a lot of work for many years and wait for a big RFP. Furthermore, they are unlikely to be able to provide everything—the soup-to-nuts infrastructure—called for in these large RFPs. Trying to do these things does not end well for start-ups. That is why the VC told the telco CEO that he killed start-ups.

The large suppliers have been reliable in providing after-sales support of hardware but are burdened with legacy product and talent portfolios that prevent software innovation. In the past, hardware-centric infrastructure components with 50-year useful lives made this long-term support very important. However, as we have seen, softwarization has changed the game fundamentally, as it has utterly changed the emphasis. As a result, these large vendors do not have the ability to provide the innovation that is required.

Many CSPs also still focus on long-term support as a key requirement. In a softwarization world, a focus on long-term support is in fact a hindrance rather than a positive. The CSPs also have many staff members that have legacy-hardware-oriented skillsets throughout all roles in the organization.

Some see open source as a solution to this problem. They think that open source software is free and that it will provide the required innovation. However, neither is the case. Current open source CSP initiatives have heavy participation and influence from the same large vendors with the same limitations. There are those in the industry who have raised serious questions about the CSP industry having the necessary scale to make open source work.  Open source by its very nature, requires a company to implement it and deliver it. In this, it is confronted with the same legacy problems described above. In the softwarized world, there is a role for open source. It has been most successful in providing components that are widely used in many industry segments and that is likely to work best for CSPs too.