I was recently in a discussion about digital architecture. There were half a dozen people in the discussion and everybody had a point of view, but there were two things prventing the discussion moving forward. The first was that we didn’t have an agreed view of digital architecture – a ‘what’. And the second was we didn’t have a ‘why’.
My knowledge of technology architecture is limited at best. You would not want me anywhere near an actual architectural design discussion. But it was the ‘why’ question that fascinated me. Why is this relevant? Why should we bother discussing this? What’s at stake here?
Let’s take a look at the competitive environments we live and work in. Increasingly, we are finding ourselves in a world of hyper-competition. What do I mean by that?
First, we are competing on new parameters. Most industry leading businesses have done the basics, so the battleground for competition has shifted. Often the differentiation itself comes from an information driven or ‘digital’ part of the service. Last month, in New Jersey, US, I agreed with a colleague that a primary basis for choosing a rental car would be the quality of the SATNAV system. But it is also likely that you have done enormous amounts of optimisation on your operational models. I just read about UPS’s optimising routing by minimising the number of left turns (in the US) and their innovation on how to land cargo planes.
Second, we are competing with entirely new types of players. Broadly there are 2 new types which require us to think harder. The first is the entry of established providers from other industries. For example, traditional banks have to worry about retailers many of whom are launching banking services. The second is the emergence of disruptive start ups. Once again, for Banks, you have new providers like http://www.simple.com or Square – each of whom may individually survive or perish, but collectively, they will pose some serious threats to your business model.
Consider two more key trends:
One, technology is now a part of the product. Almost every product. You can call it ‘smart’ or you can hide the technology but the chances that the tech component of your product is increasing annually as a percentage of your product’s value are extremely high. From Nike to Nest, everybody is doing this.
Two, there is an ongoing service-isation. As most companies seek that sweet spot between pure products and pure services, technology, and specifically software, is often the delivery mechanism for the service component. After all, when you buy the Nest thermostat, what you’re really buying is the service of ambient temperature at home. Or when you buy a washing machine from Samsung, for example, you’re really buying the service of cleaning clothes. So could Samsung theoretically also send you analyses based on weather to let you know when clothes will dry faster without a dryer? Or perhaps it might analyse the performance of your washing and let you know that you’re overloading the washing machine on weekends. You can easily think of ways in which almost every product can enhanced by a service layer, which is driven by data and software.
What does all this mean to our discussion on digital architectures? Here are some questions to ponder:
1. If your competitor provides any core service at a lower cost than yours, could you survive in the long run? If a broadcaster can run a channel at 10% lower operating costs, or lower ad-sales costs by 10%, can their competition survive? Does your architecture allow you to run core services at a competitive cost model?
2. If your competition can get new services to market faster than you can, would you retain customers?
3. How do you compete against companies that have no legacy? You can get away in an internal discussion by talking about how much of a challenge your legacy is, when it comes to making changes or adding a new service. But do your customers care, when they can get the same thing elsewhere?
4. Is your digital architecture designed to handle the new requirements of competition, new services, and the technology component of your products?
So coming back to the challenges of digital architecture, I believe there is an absolute requirement set and a competitive requirement set. The absolute set may contain things such as scalability and response time. We learnt from the early days of the web that traditional architectures and spikes on websites were not a match in heaven. Even now, major sites crash when the world flocks to see something unfold or buy something everybody covets. Architecture that can handle spikes is a critical requirement in an absolute sense. You can estimate the boundaries and plan for them. Speed/ response time is another – we did a project once where it would take a few seconds to respond, and that was unacceptable, of course. After due analysis we discovered that each request was being routed via enterprise servers across the atlantic and back, so obviously even the physical set up needs to be understood. Other absolute requirements may include security and handling specific functionality such as streaming or transactions.
But to my points above, all of this will point to navel gazing unless you also consider your competitive requirements for your architecture. This leads to benchmarking the costs of key processes, the time taken for changes, new products, upgrades and of course, your customer interface.
So whether you’re like UPS – extracting the last bit of efficiency so you can be that half a second faster than the competition – or whether you are the competition to UPS, the focus of your digital architecture needs to be (a) ensuring it can do all the things you need to do for your customers and (b) ensuring it allows you to be compete – by making you faster, cheaper, quicker or better.
That should be the starting point of the discussion for digital architecture.