Duh. That must be stupidest headline I have ever written. Of course big data is here. What I mean is there inevitably comes a time when people stop talking about something new and actually start behaving differently as a result of it. That is the moment when the ‘new’ turns into the ‘real’, or the ‘here and now’. In our line of work, what is ‘new’ often means ‘things that have not happened yet, but we expect will happen soon’. We get that all the time – why fix yesterday’s problems which (1) we have not fixed so far, (2) we do not know how to fix, and (3) is not generating the support needed in order to attempt a fix? It is easier to start talking about tomorrow’s challenges instead. To be fair, it is not silly to prepare for new challenges – you cannot prevent problems if you do not act in advance of them becoming problems. But whilst some love to talk about new challenges, some are not so keen to talk about new ways of working, and new ways of solving problems. Blacksmiths, for example, were not so keen on the invention of the automobile, and how that machine dealt with the challenge of transporting people from A to B. The rise of big data is exciting – and frightening for some – because it promises new ways of working, including new opportunities for business assurance. But the change was somewhat predictable. I think back to what my old pal Guy Howie always used to say about Moore’s Law. To Guy’s credit, in 2006 he set up his own company, and built his business model on the belief that revenue assurance practitioners had to look ahead and anticipate the ever-increasing power to exploit data. He is the kind of entrepreneur well-placed to take advantage of new technologies. To blow my own horn, Guy’s ideas prompted me to write an article describing a new paradigm for revenue assurance, which can be exclusively found in the talkRA book.
But how do we determine when the ‘new’ becomes the ‘norm’? After all, what I called a new paradigm for RA continues to be the industry exception, rather than the rule. When do things like big data stop being news and start becoming a routine fact of life for the people who really make use of it? Between 2000 and 2010, you could always find somebody willing to say that revenue assurance was new. On the other hand, a recent edit to the Wikipedia page about RA says that:
Although Revenue Assurance (sic) has always been present in the telecom parlance it has recently been brought at the forefront of the top management
RA has always been present? That sounds wrong. (And I am still arguing it should be written with a small ‘r’ and a small ‘a’ because it is a common noun, but I digress.) Some people say that RA cannot be done without relational databases but, ahem, there were telecoms networks long before there were any relational databases. And I can remember a time when you would mention ‘revenue assurance’ down the telco canteen and the rest of the table would give you a confused look, because they had no idea what you were talking about. The problem with dating and describing trends within a global industry is that developments occur at different times in different companies in different parts of the world. For example, mobile money might be new in many parts of the world, but it is no longer new in Kenya. Consumption of mobile data might be ‘exploding’ now, but the Japanese had i-mode when many Americans still thought the Motorola StarTAC was a pretty cool phone. The same phenomenon is going to occur with big data (which even though it is very very big, should definitely not be written as ‘Big Data’). Some businesses and people will have already begun the process of adapting to big data, whilst others will keep talking about its newness. So how should we judge when the new becomes mainstream?
They say that money talks, and the rest had better take a hike. Applying this maxim to language and business, a word or phrase becomes mainstream when somebody intends to make money by using that word or phrase to attract a mainstream crowd who want to talk and hear about it. In other words, a word becomes mainstream when somebody arranges a conference which is described using that word. Another old pal, Hugh Roberts, started pulling together programs for the first RA conferences in the early 00’s. So contrary to what it now says in Wikipedia, I will use that as my personal guide for when ‘revenue assurance’ entered the telecom parlance – though if people know of earlier examples, I would be very happy to be corrected. By the same measure, I must tip my hat to my relatively new pal, Dan Baker, who is leading the way when it comes to promoting a new mainstream that will combine big data with business assurance. Dan is hosting a workshop about financial and revenue analytics to be given on January 29th at Telecoms Analytics 2013 in Atlanta. (I wish I could be there to join him, but anyone planning to be state-side in January should know that followers of Black Swan can obtain a discount for the event.) The conference is billed with the intriguing tagline:
Big Data is the Tool Set.
Real-Time Analytics is the Objective.
Does that mean that big data is already part of the tool set for the average business assurance practitioner? Hardly. Many of them will be unaware that their business could achieve a quantum leap in assurance capabilities through deploying technologies like Apache Hadoop. For the foreseeable future, the vast majority of business assurance practitioners will continue to derive their automated analytics from queries of a relational database that contains only a tiny subset of the company’s total data, and which their department uniquely ‘controls’. In fact, I think the use of the word ‘analytics’ in the conference title will encourage some people to attempt to deliberately confuse the audience by asserting that relational databases can provide the same capability as big data. Analytics is anything but new, so the exciting part of this event relates to what can be analysed, which is being greatly extended because of big data technologies.
Remember that it was only in 2008 that a certain CTO made the biased argument that RA departments must only use dedicated databases that have unique feeds of raw data. His rationale was that there is always an unacceptable risk of error if the data being analysed has been loaded to a ‘secondary’ database instead of one dedicated to RA. He said managing a common corporate-wide source of data for multiple analytical goals was a wrong-headed way to cut costs, and he justified this by insisting there would always be unforeseen additional costs because the analytical results would all be wrong! Well move over, old-timer, because big data is going to make a mockery of the idea that each department in a company must do its own special analytics on its own special relational database with its own special data in it. That approach is great if you want to maximize the profits by selling duplicate systems to a telco, but it is an inefficient way to run a telco. Thankfully, the sheer scale of big data will grind down the silly and parochial arguments about who controls what database within a single business.
Big data is a genuinely transformative technology. It will yield an incredible opportunity for guys who started out performing CDR reconciliations to massively expand their insight. They will get the chance to use this increased analytical power to tear down traditional obstacles and find multiple new ways to boost the bottom line. But they have to take their chance, not wait for someone else to show them how. Practitioners with imagination will embrace the possibilities, acting as entrepreneurs within their own company. It is also good to see Dan in the vanguard, and it is encouraging to see that the connection is being made between big data and business assurance. But just like the early days of revenue assurance, you can be sure that some people will talk a lot of rubbish about how to lever big data to gain new perspectives on the business, when really they just want to muddy the waters as much as possible. And for every day of the next decade, you can also be sure that someone somewhere will be telling an audience that the use of big data for business assurance is completely new, because it was only discovered by them that morning. Then, in the year 2024, somebody will edit Wikipedia and we will read that it is impossible to do business assurance without big data, which leads to no contradiction, because Wikipedia will also say there was never a time before big data. But none of that nonsense really matters, because all that really matters is that the good practitioners know that the true power of a tool is demonstrated in use, and they will set the real standards when they make use of big data.