One company that I've come across has come up with a pretty good basis for a definition of digital in their marketing literature. Cognizant describes digital as "SMAC" — i.e., social mobile, analytics, and cloud.
However, to get a little way away from marketing spiel, and to add some extra bits, let's take SMAC as a root and extend it. To those four we'll add consumerisation of IT, and a specific mention of big data. Together, those six things should form a decent basis for describing what digital is when it comes to the commercial realities of how to introduce it into the business.
When we think about social, the key is to consider why social is happening, rather than think of it as just a set of tools. For example, Facebook, Twitter, and so on are tools, but why people use them is much more important. The same was true with the internet when we first started using that — that was a tool, but what it did to the lives of normal people in terms of access to information, increased freedom, etc., was much more important.
A digital business uses social to interact with customers, both proactively and responsively. Without that meaningful interaction, social is pointless. Likewise, social must have specific SMART goals and objectives. That's particularly something to watch as it's common to find businesses undertaking social strategies without backing efforts up with proper measurement.
Another aspect that gets missed is user experience. When the company runs campaigns on social media, it needs to be joined up with good consideration of the customer journey and their experience. That means it typically is a multidisciplinary effort with specialised UX support, rather than just being something that happens in isolation within the marketing team.
Mobile is a similar shape to social in that it's the why as to why people use mobile devices as opposed to anything structural about the devices themselves.
I've written about post-PC frequently on these pages, but the easiest way to understand the appeal of mobile is that people tend to like how it is non-intrusive and non-interruptive. It allows people to get on with their real lives, and easily blend in aspects of their digital lives without hassle. It's much easier to pick up a mobile device when you're out and about and shoot off a quick email to someone than to remember you need to do it when you get home, then get home, then log on, then do it, and so on.
A good way to look at this is what social and mobile tend to converge into a pattern of "non-voice customer service" -- i.e., how an individual can use aspects of their digital life to interact with a business digitally, rather than have that interaction impinge on their real lives.
Analytics and big data
Big data isn't one of those buzzwords that I like because most people get the general principle wrong.
The idea behind big data is that you can derive understanding about behaviour through statistical analysis of clumps of data. You can then take that understanding and implement some form of control to either get more of what you want, or get less of what you don't want.
The big problem with big data is that statistical analysis tends to be very dumb. Determining behaviour is one thing, determining motive behind behaviour is quite another. You may know that most women buy from your commerce site on a Friday evening for example, but to really take advantage of that fact you need to know why that happens.
Again, this comes back to question that is about "why." It's not enough to just know "that something happens."
Consumeration of IT
This one, and the next one on cloud, is primarily about a change in how companies buy technology.
As tech companies look to ensnare normal consumers, the products they produce get better (as customers are more fickle than businesses), and cheaper (the market is bigger, so economies of scale apply). Businesses can take advantage of both of these facts, and as a result, IT no longer needs to be specialised.
A good example here is BlackBerry. Previously getting work email onto a phone was relatively painful and involved specialised hardware and software, and specific airtime contracts. It no longer does with BYOD and CYOD (choose your own device) removing that specialisation.
Another dimension to consider here is that as consumers — who are also employees — get more exposure to apps, and their attendant services, they get better at innovation. Although companies generally are either wary of or downright hate employees using Dropbox to get data into and out of the business, the act of doing that is innovative.
Finally, we come to cloud. As mentioned, this is really about how companies buy. There are all sorts of reasons to like outsourcing IT functions to the cloud, whether it's just outsourcing compute power into a load of servers that you run as if they were your own, or buying functionality on an SaaS basis.
Is cloud necessary for digital? To an extent, it likely does not. However, as a fashion/trend, it's clearly important, and a lot of the tools and services involved in digital are unlocked as part of a cloud-based approach, hence it's likely important.
The important thing here is that "digital" means something specific, and reasonably clear. It's not an opaque buzzword.
Like post-PC, it's a sociological change, rather than a technical one. You can see that by the fact that this is generally all about the "why" this is happening — why are customers using social, why are they using mobile, why big data is showing the trends that it is, why are companies able to buy and use consumer products, and why is running systems in the cloud easier.
To understand digital, you have to understand the "why."