There was lots of conversation last weekend about Verizon’s latest entry into the cloud market. I say latest because their purchase of Terremark means that they have indirectly been in the cloud business for quite some time. Despite the commentators saying Verizon’s offering is what OpenStack should have been (I don’t agree, but that’s a different subject), Verizon appears, at least from the marketing fluff, to have done some interesting stuff.
Interesting not because of the bold technology claims being made, but because they have gone for differentiation from the competition by engineering something which seems truly different- something which may sound obvious but would be a revelation in a telco. Whether their technical claims are true or not is not really relevant here, what is relevant is that they may have executed, if indeed it is true, with the aid of inorganic influence (the Terremark acquisition), an incredible cultural shift.
Verizon’s heritage is as a telco, a provider of pipes. They don’t have a traditional heritage in IT services, managed services or cloud services. The culture that the telco world has, at least as it exists today, is very different from the IT service world.
In the telco world your products are not measured on differentiation from a technical perspective but instead are, as long as you aren’t running ancient technology that can’t provide a modern feature-set, measured by reliability, geographic reach, the bandwidth options available, service delivery reputation, price, and customer service.
Successful and even innovative products in the telco world don’t require any super clever engineering by the telco (at the risk of the wrath of my colleagues in this space) as you aren’t exposed to the technology at the level where you can genuinely influence how it works – all that happens in the vendor’s engineering function. There is a lot of plugging things together and making it work (which isn’t easy I hasten to add to lessen the beating). It requires a lot of capital investment. It also requires working with vendors to make sure their products can work together the way you want.
As an IT service provider you are currently measured on the same qualities with the added measures of genuine technical differentiation and innovation. This isn’t to say these are necessarily explicit tick-boxes on a customer’s vendor evaluation. Explicit tick-boxes do exist though in meeting customer expectations of a service level guarantee provided by their suppliers. You can’t just pick standard parts off the shelf and plug them together, because for the most part the plugging together really isn’t that simple. You actually need to do some deep engineering to make these service level guarantees a reality, most likely amassing IP along the way.
Arguably this is the way the telco world was before the standardisation of many of the underlying technologies. The telco way may be what the IT services way becomes when there is little differentiation to be had through technology.
Now imagine the cultural shift that needs to happen after two decades or more of operating in the telco way. Building something rather than productising something from a vendor is not something that telcos are good at (anymore), it’s just not how they are organised to roll.
As a telco, you want to build a new service? Great! Go spend time with the Ciscos and Junipers of the world and work out how their products will play nice your existing infrastructure. Spend time working out how you will operationally support the ‘new’ product. Build some marketing collateral. Sell it to the market.
The formula is relatively simple and the challenge becomes more around how you keep up with migrations as the technology goes through its various evolutions, and how you squeeze every operational penny from the machine to fight price erosion and increase margin.
Over time this market has become more and more commodotised which is why telcos now want in on the highly lucrative IT services space. For some reason telcos seem to naively think that being the provider of the pipes puts them in pole position when it comes to the customer’s selection of IT services. That’s like my electricity supplier thinking that it is in pole position to sell me a fridge- it really has no such advantage.
Most telcos, however, don’t want to leave the comfortable world of slapping their own brand on someone else’s product whilst passing it off as their own. This engineering stuff is difficult and expensive in comparison. Telcos that do it the telco way are, in my opinion, going to fail due to not showing any differentiation between themselves and others doing the same, further driving down the prices that they can sell at, reducing the very margin they seek to find through their foray into this new world- differentiation is what makes engineering so important.
The problem with engineering is that it takes significant (and continuous) investment requiring construction of a feature-set from pretty much the ground up when there are products on the market which will give you a leg up. When you are trying to make an impact on a market where you are already on the back foot in an area where the market leaders have a better understanding of the market, a business case describing a significant and ongoing engineering effort to support market entry can be very difficult to make. This is more true when you don’t have an engineering organisation in the first place as standing one up takes time (and money) and puts you even further on the back foot. As a telco, if you are to go down this road, you really have to go all in. Otherwise you’ll just find out that the initial leg up quickly turns into a barrier to making the really cool stuff happen.
Another part of this conundrum that telcos need to become very comfortable with, and very quickly, is the idea of rapid iteration. Typically, telco product development cycles are between a year and 18 months in duration because the underlying technology doesn’t usually change in that timeframe. This means satisfying a large and comprehensive feature-set in a single (and large) expensive development effort.
This is very different from the IT space where there can be a technology shift within a matter of months. To counteract this, and to keep the feature-set current the provider must rapidly iterate on its solution, making lots of smaller steps to achieve the desired feature-set so as to demonstrate progress to the market whilst at the same time not falling behind its competition. This requires a never-ending effort, one that can never have a line drawn under it marking it as complete.
As a telco, introducing these principles will, in my opinion’ result in one of two approaches. Either the telco divides itself into two with one half operating in the telco way whilst the other operates in the IT service provider way, or the telco evolves its methods to move the telco half to work the IT services way.
With concepts such as SDN and network functions virtualisation, I think we’ll begin to see the same rapid capability evolution in the telco space as we see in the IT space. Programmable pipes being leveraged for several previously unthought of use-cases will mean that telcos will need to start engineering and not just installing the latest iteration of a vendors product. This will require a widespread adoption of the IT services way.
Going back to Verizon it will be interesting to see, if they have indeed made this shift, if and how they will apply these principles to their traditional telco business. Interesting times lay ahead for both Verizon and other telcos as they try to maintain their relevance and profitability.