Big Data. BIG Data. Big Big Big Data. There are some arguments about how to define Big Data, but I think the name pretty much explains itself. It is data, and it is big. What makes Big Data interesting is that technology now allows us to scale up all the things we always did with small data. As well as looking up and creating new data records, like books on a library shelf, we can also aggregate across the data, searching through it all, performing computations over the whole data set, and so on. When you do those things with Big Data, you see things you could never see with small data. At a basic statistical level, what might have looked like an outlier or a quirky one-off result within a small data set can be seen to be part of an important recurring pattern in a much larger data set. And there is value to be made by seeing patterns. Big science, like CERN, generates scientific value by discovering patterns in the huge volumes of data that captures collisions between incredibly small particles. Big politics, like Barack Obama’s Presidential campaign, generates political value by calculating the most cost effective TV ad buys, in terms of precisely maximizing the number of people who are influenced to vote for their man. Naturally, big business sees an opportunity to make value from Big Data too. And yet…
Last week I attended a very enjoyable dinner for members of the press, hosted by WeDo, the Portuguese vendors of business assurance software. I am not sure why WeDo invited me to a press dinner – as a blogger, I consider myself an imposter, whenever sat alongside the heavyweight journalists of the trade press. However, I was glad to be there, as the conversation about Big Data was sharp, incisive, and witty. The Chatham House Rule applied, so I cannot tell you specifics of who said what. However, the overall theme was clear: there was agreement that Big Data is going to make a lot of money for some companies. Just not for telcos.
Nobody doubts that telcos have a lot of data – but note the distinction between having a lot of data, and the concept of Big Data, which also requires the capacity to do things with that data. And there is no doubt that a lot can be learned from telco data. Just analysing large volumes of billing records is enough to yield all sorts of useful insights about an individual customer, as recently vouched for by the NSA. But the brightest and best of the trade press, as assembled by WeDo, who also know plenty about finding value in data, could not make themselves believe that telcos will make the entrepreneurial leap forward, or lateral step sideways, that is needed to make a lot of money by exploiting Big Data. Telcos are surely competent to implement Big Data technologies. Their doubts revolved around whether, once the Big Data technology was in place, telcos would know what to do with it.
I was reminded of a previous headache for the telecoms industry, that was also biggish and involved data. Do you remember the hunt for the killer app? In that instance, customers were going to get much bigger bitpipes which would allow lots of data to come sloshing through to their handsets. The challenge was working out what people might want to do with it. Technology had developed at a rate that outstripped our understanding of human motivation and desires. A similar problem occurs with Big Data. We know the end goal that businesses pursue – they want to make money. But handing a telco the keys to Big Data, and telling them to make money from it, is a bit like handing someone the keys to an oil tanker, and telling them to make money from it. Clearly some people make money from oil tankers, but you cannot just hand the key to anybody, and expect them to know what to do with it. For all the big talk about Big Data in telecoms, some of the best informed people in the industry have not yet come across real examples of telcos making money from Big Data. And around the table, there were too few examples of telcos who even had good ideas about how they might exploit Big Data.
I admit that I was stumped too. During the course of the conversation, it occurred to me that I had heard a fair few presentations and sales pitches relating to Big Data, but none of them had pushed a convincing ‘killer app’ that would make lots of money for telcos. Despite my blogger/imposter status, I almost feel like I should make suggestions. But that is not the point. Nobody knows for sure what will make money, until the data is scrutinized. What we need is not a killer app, but killer questions. And we will only discover the killer question, after it has been asked. We will recognize the killer question in hindsight, based on the answers yielded by the data.
There is Big Data in science, and telcos need to learn from science about how to construct a productive methodology, followed in a disciplined fashion, designed to process vast amounts of data in order to deliver valuable new insights. The funny thing about that, is that it also means recognizing the importance of imagination, in following that methodology. The combination involves tools and technology, but also the human vision and intuition to construct hypotheses about what the world is like. Sharp new ways of looking at the world are a crucial input to the process. If corroborated by experimental data, scientific hypotheses can yield revolutionary new opportunities. I have full confidence that telcos have the skills to manage the data. As the participants at WeDo’s dinner agreed, we have less confidence that telcos know how to construct radical new hypotheses for making money, or even if they want to.
From personal experience, I have dealt with telco people who talk about innovation, but rarely do anything innovative. I do not consider that to be a significant failing. Behaving like a utility, and trying to keep costs low whilst earning a steady margin from a service that people will always want, during bad times and good, is a perfectly valid business strategy. It may very well be the business strategy that investors most want telcos to pursue. Telcos are definitely migrating from a high growth, high P/E ratio proposition – the perceived norm for the world of tech – toward better appreciation of their defensive qualities as a stock, with greater focus on dividend yields. This befits big businesses that provide utility-type services, in markets with significant barriers to entry and hence subject to extensive regulation. So let us be clear: there is nothing inherently wrong with the idea of a business that is not very innovative.
Furthermore, the whole idea that people like innovation should be approached with some scepticism. This excellent academic paper, from psychologists Jennifer Mueller, Shimul Melwani and Jack Goncalo, describes studies that tested bias against creative ideas. Whilst the participants in the study felt themselves to be keen on creativity, and in favour of creative new solutions to established problems, they were actually much more risk averse in their choices than would be consistent with their conscious beliefs about themselves. I like to think of this as scientific corroboration of the folk wisdom encapsulated by the phrase: ‘nobody ever got fired by buying IBM’. Whilst people like to think of themselves as open to change and improvement, in truth they are more motivated by the fear of taking a risk that backfires. And so large corporate entities, like telcos, have reason to be stacked full of people who talk positively about innovation, but never do anything remotely innovative.
Big Data will make most money when applied innovatively. And yet, the size of Big Data tends to limit the possibilities for small, entrepreneurial, risk-loving firms to take the chances that will promote big leaps forward. Or maybe those smaller firms will succeed, but only in the growing sphere of government data that is increasingly being opened up, at some taxpayer expense, to the general public. Here, I think, is an example that telcos could learn from. In this unusual instance, big government might be better at promoting win-win relations with entrepreneurs than big business tends to be. By opening up the data vaults, governments stimulate small, unpredictable, and agile thinkers to find new ways to make value from data.
Telcos will not want to just give their data away. But there must be ways they can seek risk-reward partnerships with a broad range of entrepreneurs. After all, this actually happens in telcos all the time. Unfortunately, it tends to happen at the level of a telco employee, with insider knowledge, leaving the telco in order to better exploit their inside knowledge. The employees leave, take all the remaining risks, but they also keep all the rewards. Telcos receive zero benefit when this happens. It would be healthier for telcos to find ways to transition individuals from the payroll/employee model to more of a risk-reward/partnership model. The aim is to keep the brighter members of their team in the orbit of the telco, and motivated to create value. This is better than stifling their creative impulses, by failing to reward them. And if telcos cannot find ways to partner with the human capital they already have, who already have a head start in understanding the business, its customers, and its data, than it will be even harder for telcos to recruit, and retain, bright and expensive people from outside the company.
Past experience makes me wonder if telcos could ever realize such a Google-like paradigm, where employees can also be business partners. To cite another example from the history of telecoms, I always thought telcos were making a bad mistake, and indulging the worst kind of closed thinking, by creating walled gardens of the internet. Customers were inevitably going to demand the whole thing, sooner or later, so the money spent on making walled gardens was mostly wasted. That phase in the maturation of telco business models revealed a conservatism that was counter-productive. But telcos are older now. People get more conservative as they grow old, which hardly encourages optimism that telcos will find ways to motivate the best, most talented, most imaginative brains to unlock their data, to discover new sources of value within.
But maybe I am wrong. This is just my hypothesis. Like Deming said: in God we trust, all others bring data. I look forward to seeing the data about how telcos exploit their Big Data.