In general terms, we believe strongly in the principle of continuous innovation. Companies such as telcos need to continue to innovate to stay relevant and to ensure that they are not diminished in the perception of their customers. The problem is, of course, that telcos have struggled to innovate in so many ways in recent years. And yet still they need to offer more and more, just to maintain their position, like the Red Queen. The question is, if telcos have been so singularly unsuccessful in recent years in delivering applications to consumers (beyond core services like voice, messaging and access), where will that innovation come from.

The standard answer is third parties in the Internet world. Open API access, in principle, offers a means to outsource this innovation to a willing community of developers. However, this development community has taken on an almost mythical status in recent years and telco-led initiatives to attract developers seem to have achieved little. At the same time, the Internet world has managed to find this community and continues to tap into it to fuel innovation. There are now more than 6000 APIs, at least according to industry tracker Programmable Web. Of these, more than 5000 are based on either REST or SOAP. Clearly, developers don't much care about the specific API messages; they care about the general model. These have largely emerged in the past three or four years – a period during which the telco industry has struggled to launch OneAPI.

But there are signs that things are changing. Dimitri Sirota from Layer 7 provided some great insight into why their telco customers are starting to offer access to SOAP or RESTful APIs via gateway solutions that aggregate and federate access to resources in networks. The interesting thing is that this a) shifts the burden for new application development towards a community with somewhat better credentials to understand consumer appetites for applications; and b) enables an environment in which the cost of failure is reduced.

This last point is instructive. If the cost of failure is high, the bar is set such that there are disincentives to actually launch new services – or, not to launch them until they have been perfected (thereby increasing development costs significantly and deferring any potential payback). But we know from the Internet world that services don't have to be perfect to be successful. Customers are interested in applications and services that offer utility and are prepared to accept a trade-off between utility and completeness. The expectation is that, if the service offers some value, it will improve and, ultimately deliver more.

There's a famous passage in "On the Origin of Species" in which Darwin discusses the utility of an imperfect eye. He explains that, some degree of light receptivity is better than none and that each incremental improvement can contribute to the evolution of better sight reception and hence to a more advanced eye.

It's a good argument, but it also has lessons for us here. We may have a particular service goal in mind, but we don't necessarily need to be able to offer that service right from the start. Services can evolve, step by step – and based on real customer interaction (acting as competitive pressures and selecting positively or negatively for services that are useful). What we need is an adaptive environment for innovation.

Such an environment consists of:

• API exposure to telco capabilities;
• Incentives to use them (developers care about value and utility, not about the specific format of an API);
• Feedback mechanisms from a customer base; and
• Low risks of failure (if something really doesn't work, throw it away and try it again).

With such conditions, telcos can enable innovation and enjoy the benefits of enhancing their network with richer services, regardless of whether they are offered directly, or indirectly via the development community.

In such an environment, half an application that does something is better than talking for years about how to enable such applications but doing nothing. 10% of an application that delivers some utility is better than nothing. Darwin was right: anything that confers a positive benefit can be better than nothing. And with the right developers acting as instruments of evolutionary change, applications can evolve and be improved. Failing applications will die, but die with less cost and fallout than ones that have been perfected and honed at huge cost.

And that's why we are excited about TMW and recent developments in the industry. Telcos may just be finally realising that they don't need to standardise everything, that they can go it alone and that there are relatively simple ways to reach out to developers who can enhance their network and leverage their assets. They may also be recognising that their assets are diverse and not necessarily what they thought – i.e. it's not just about voice and location any more, if it ever was.

All they need to do is make it worthwhile (don't try to take too much rent) and provide the infrastructure (deliver a platform). We have no idea what constitutes a successful service – who really does? But by leveraging technology such as that offered by Layer 7 and others, telcos can start out on a process of fostering and promoting evolution, which will, hopefully, lead to a flowering of innovation and adaptation.