Currently showing: Other

24 Jan 14 07:57

It seems that everyone at the WEF is talking about data this year.

Technology has revolutionized the way we do business, but also our everyday lives. In developed countries, consumers increasingly shop, communicate, bank, lobby.. even date! using digital technologies. In fact, in many ways, consumer technology use is well ahead, and often drives, business technology use. In developing countries, the shift is sometimes happening even faster. Countries like Kenya (which leads the world in mobile money) are fast-tracking straight to mobile technologies, to fuel growth:

Even businesses like Wal-mart are becoming a technology company -amazing! In its relentless drive to cut prices, Wal-Mart embraced technology to become an innovator in the way its stores track inventory and restock their shelves, cutting costs and passing the savings along to customers. For them, optimizing their supply chain has been key to their success and requires significant technology and analytical skills.

So what does this mean for jobs? During these first few days at this year's WEF I must have met at least half a dozen data scientists. I see them constantly surrounded by business people trying to understand how we can do more with big and unstructured data.

The challenge is sorting through these vast amounts of data to find what is useful and can give companies insights into consumers and competitors so as to develop a competitive edge. But sorting through data will only provide useful insights when one asks the right questions, which is not always easy. Without focus, or with the wrong focus, businesses could incur significant economic losses by using the wrong data.

In my view, this simply requires companies to re-shift their business approach. And the sooner the better as it's here to stay. We are already seeing winners and losers coming out of this.

What are your views? Is big data just a fad? Who are the real winners? Which companies are ahead, which are behind?

Category: Other

Location: Davos, Switzerland


Alicia Montoya - 24 Jan 2014, 4:53 p.m.

Funny, I just came across this new KPMG study that found that while C-suite executives today are striving to drive data-centric transformations of their businesses, most are struggling to connect the dots.

Interestingly, even KPMG is adapting: The report was made by KPMG Capital, a global investment fund created by KPMG International in November to invest in innovation in data and analytics (D&A). One of the winners? We shall see!

Here's the story:

Andrea Ferrario - 24 Jan 2014, 5:30 p.m.

This is an interesting article. I would see "Big Data" as a highly strategic holistic business opportunity; "strategic" because it requires structural changes in the business model of whole sectors of a company, "holistic" because it is transversal to the classical subdivision into well-defined business units. I believe that the Big Data opportunity can be characterized by

-Visualization & predictivity
-Time to market

Personalization refers to the necessity of designing a strategy which takes care in toto of the business model and architecture of the company. It is a complicated mixture of strategy and highly sophisticated methodologies. For example, in reinsurance I would start with an exhaustive inventory of the currently used quantitative models to quantify and to assess model risk running through the whole company. The big review & validation processes really lay the foundation of a taylor-made and solid Big Data implementation! How? Well, through the identification of sources of "unconventional/ unstructured" data, upgrades and new mathematical frameworks, cloud solutions etc... that can dramatically boost performance of the validated solution. In summary, there exists no "general Big Data solution/framework"; everything really depends on the company under exam and its business needs.

Visualization and Predictivity are at the core of the business empowerment through Big Data solutions: Visualization condenses the hidden value of highly dimensional data into the space time framework perceived by human mind; it is a great tool for modern business intelligence. Predictivity is the result of the state of the art (most of the time proprietary) solutions which win the quantitative side of the Big Data challenge: of course, it is greatly supported by Visualization.
Both aspects require an extensive amount of skilled resources: being able to properly visualize information to steer decision making and obtain the best predictions is definitely one of the keys to success in the Big Data challenge and probably the most difficult one to find.

Optimization (of processes, supply chains...) and Time to Market (products & services) are the final product of a sound Big Data strategy.
I see the Big Data challenge really as a huge opportunity, once the business needs and targets have been properly defined.
Being able to talk the language of IT, business intelligence, management and synergizing resources in such complex and innovative projects
is what makes the challenge so interesting and rewarding.

Marty Ellingsworth - 24 Jan 2014, 9:06 p.m.

Jayne is spot on. I've seen many company and carrier teams struggling with core systems reporting and analysis who are now trying to answer their Board and C-Suite on Big Data questions.

Consider plotting your progress on the template in this article

and then make your roadmap for next steps part of your strategy process for investments in people, capabilities, data, and analytics.

Dan Sullivan - 27 Jan 2014, 7:14 p.m.

A key point here is the need to sort through and analyze large volumes of data. In many ways collecting and storing big data is the "easy" part (at least relative to analysis). Analysis should be hypothesis driven with a focus on changing a business process. It's good to know how to segment your customers but is is better to know how to change your relationship to those customers to increase sales, improve retention, etc.

We need to stay focused in big data analysis because there are so many questions we could ask. The best methodologies for big data analysis will look a lot like the scientific method: they will force us to articulate a testable hypothesis and then analyze data to support or disprove the hypothesis. Ideally, if our hypothesis is false, we will fail early saving time and money.

Oliver Werneyer - 28 Jan 2014, 11:44 a.m.

Hi Jayne

Great article and perfect timing. I recently did an interview with a company on this topic (

I agree that just having data and the technology to analyse it is not enough. There are several other component that will determine the usefulness and quality of the insights of which the main ones for me are:
- The human element: Especially in insurance, remember that in the end a human being will be buying the policy and providing cover for them. If we machine the insights to much we might not have done the right thing for the policyholder
- Don't lose sight of your goals: I have heard many people talk about analysing a database and just see what comes out. I think that is dangerous and wastes a lot of time and money. If you do not know why you are doing this or what you are looking for then you will get caught up in irrelevant, wrong and out-of-date insights and waste time and money on things which have little or no impact on your business.
- There are implications: Keep in mind that from a data protection and personal rights perspective the market still has to catch up quite a bit. There will still be some major changes. Be careful not to build things which you know actually infringes on people's rights to privacy but you do it because it is still legal. What happens when it changes. What happens when your expensive and integrated system n longer has access to what it needs to work? Money down the toilet.

Jennifer Rodney - 28 Jan 2014, 12:35 p.m.

Thanks for sharing Oliver - interesting article on

I think there are times when not having clear goals can be helpful to data analysis. Going in with pre-conceived notions can cloud interpretation or application of the numbers when leading questions are the only ones being asked.

Being open to really just see what's there (or what isn't - despite one's hopes), one can be surprised, learn unexpected things, sometimes even face unwelcome but healthy truths about the success of their product/project/campaign/etc.

That being said, the quality and relevance of the data being examined is of paramount importance!

And you raise an interesting point about the impact data protection might have in the future - hadn't thought about it from that angle.

Oliver Werneyer - 28 Jan 2014, 4:14 p.m.

Hi Jennifer

I am with you on that one, you can definitely find out really interesting stuff when keeping an open mind. But, that is not what I tried to say. I would avoid getting a whole bunch of data and say "So, let's see what we can get out of this." without already knowing at least one or two things you want to establish. There was a nice example in the book "Freakonomics" (as I remember it). They showed, through statistical analysis, that there is a correlation between the number of pizza slices sold in a certain area and the amount of crime committed by youths between ages of 16 and 21. Now, purely from a statistical perspective there is a correlation, but does it make sense to consider this? So, once pizza slice sales see an increase then put more police on the street monitoring 16 to 21 year olds?

I would imagine that you would collect a bunch of data to test some theory or get a view on something. At this point you should be open to see other interactions and results you might have not expected and then drill down on these in a follow up session, applying reasonableness checks and framing the data and questions correctly. I just think the value of the insights will be much better.

So, keeping an open mind, YES. Go in blind and wait and see what you can get, that gets dangerous (in my view). You could get the weirdest and most useless insights and because you can not properly frame or place the insight, might end up using erroneous insights.

Yes, regulation is such a tricky subject at this moment. It is so far behind and will, at some point, leap quite significantly in order to catch up.

Bilal Zafar - 30 Jan 2014, 6:46 p.m.

Jayne, it's interesting that you mentioned Walmart. Initially it may seem that retail and technology are completely different worlds, but if you dig deeper, big supermarket chains are extremely well positioned to capture – and use – behavioural data on customers. Predicting consumer behaviour is, of course, the holy grail of today's business world and big data can help achieve that.

Tesco is a great example of this. They introduced Tesco Clubcard (a loyalty card scheme) in the UK in early 90s. They now have 15 million Clubcard holders. They are in an excellent position to run data analytics on customers (they know the name, address, age etc. when the person signs up) against what they buy.

Tesco's then-Chairman Lord MacLaurin, famously said to the marketing company that introduced Clubcard "What scares me about this is that you know more about my customers after three months than I know after 30 years."

One can imagine that by analysing what customers buy and how often, they can predict things like a birth in the family (purchase of child products), affluence (do they buy budget or top of the range?), and how often and where they travel (gaps in store visit, visit a store somewhere else in the country) and the list goes on.

This is also relevant for insurance of course. It doesn't come as a surprise that Tesco has recently set up an insurance company in the UK. Does the data on 15 million customers (20% of UK population) give them a competitive advantage in the tough market that is UK insurance? Absolutely, but only time will tell if they are able to use it to their advantage.

This shows however that everything is interconnected in today's world, from supermarkets to travel companies, to insurance and everything else. Using the data within an industry can be very powerful, but the true power is unleashed when the dots are connected.

Paritosh - 31 Jan 2014, 4:10 a.m.

Thought 1 - What if I told you that big data and what it could do, was clearly thought about science around 1940s??? It still lists as fictional science in wikipedia. Psychohistory. One reason why Isaac Asimov is respected in the world. And this is another reason why one should read old literature from various countries before coming to monk who sold his Ferrari and likes.

Thought 2 - The very important point you made here is on how to use this data will have the key. So an engineer equipped with how to skim the data, will find something that he think could be useful for his company (e.g. a reinsurance company), a doctor will find his way, and even a "Scriptwriter" will find his way to the same logical conculsion. A company would certainly gain if all three of these and many other from "diverse" fields are on the job with a single goal of finding something useful to the company.

Peter Münzenmayer - 7 Feb 2014, 7:30 a.m.

An interesting report indeed, I like the last sentence most, "Throwing all available data into a pot and hoping for a tasty stew of insights will rarely — if ever—deliver meaningful results."
After last years hype around the topic I see some light shimmering through the fog separating business value from technology advancements. I believe we will see more concrete applications and true business value in 2014 making the topic more tangible.

Peter Münzenmayer - 7 Feb 2014, 7:45 a.m.

Hi Jennifer and Oliver
A great discussion. Let us move the thoughts further, no matter if we use a data-driven or a hypothesis-driven approach the most important question is ... what do we do with this insight? A colleague once told me that information has no value... a very provocative statement, but if you think about it, as long as you are not able to derive any actions or decisions it will at least add no economical value.
In this respect no matter how you derive insights from data as long as you don't follow through, verify your hypothesis and make something of value with it it is probably waste of time and money...

Andrea Ferrario - 8 Feb 2014, 9:42 p.m.

I believe information has value by itself as it is the final result of the process of giving form to ideas: to multiply this intrinsic value is the real deal in business, though. Having insights is already a good starting point! I find this point extremely nontrivial. In my experience, even insights from data or working assumptions are a (starting) result by their own. As you suggested, one should test them in a simplified framework driven by clear business needs...otherwise the whole discovery process stops or gets out of control devouring time and resources.

Peter Münzenmayer - 9 Feb 2014, 12:14 p.m.

Dear Andrea
I like your thought and you are right, information can lead to insight and eventually to better informed decisions. Cross-functional insights can be strong sources of innovation! But unfortunately there is a huge information overload and chasing information without a purpose and even if it is only for self interest can become dangerous.
The right balance between business and personal needs is the goal.

Andrea Ferrario - 9 Feb 2014, 1:45 p.m.

Dear Peter
thank you for your interesting reply. You are right: there exists a tangible risk to be overwhelmed by the sheer amount of data and -for example- trying to apply general frameworks to peculiar business needs. One should try to identify and solve reduced / toy problems in a specified business area, instead. I see this as a first joint effort of management, BI, IT harness Big Data.

Jennifer Rodney - 11 Feb 2014, 11:59 a.m.

Peter and Andrea, I think you both make great points. There's one more element I would add to the mix - which is a corporate culture that supports innovation and changed based on the results of learnings garnered from all that data.

From my own professional experience, I've seen data analysis result in simple but effective cost-reducing production adjustments in my last job. These changes were put into place within months in part because of a non-hierarchical, straight forward environment I worked in.

In my opinion, the willingness to act on findings and to empower those closest to the data to make recommendations or better yet changes can help make big data practically useful.

Andrea Ferrario - 11 Feb 2014, 12:11 p.m.

Jennifer, you move the Big Data challenge to the "metalanguage" level: this is an interesting consideration. Not everybody can speak all different languages spoken in a complex business environment...this ability makes a huge difference in the final application of results, imho.

Alan Brasunas - 10 Mar 2014, 6 p.m.

Hello Jayne,

Great posting and some very insightful comments generated.

While I agree it is necessary for companies to shift their business focus or way of doing business, this is not always so simple depending upon the culture of the company. Steve Blank, a retired Silicon Valley serial entrepreneur, recently published an article titled "Why Companies Stop Innovating," which sites some of the impediments hindering companies from embracing change:

While applications for Big Data may only be limited by one's imagination, Ron Kennett suggests an interesting one: Managing Black Swans. A recent article published on MIT Sloan Management Review titled "The Science of Managing Black Swans" also provides a link to the abstract of his recent paper:

Ngan, Sophia Van - 9 May 2014, 4:14 a.m.

Hi, I’m reading about Big Data and I found this page. I thought your discussions are very interesting, so I’d like to stop by to share my thoughts.
Success of Big Data and Smart Analytics in an organization is dependent on so many factors, which pose many organizations to real challenges of execution.
“Big data is like teenage s.x: everyone talks about it, nobody really knows how to do it, and everyone thinks everyone else is doing it, so everyone claims they are doing it…”
This message was posted on the personal Facebook page of Dan Ariely, the best-selling author and Professor of Behaviour Economics at Duke University. It has been shared nearly 800 times across the social network, liked by over 1,700 people, and widely quoted and discussed in many blogs, sites, and forums. It has been instantly a hit because it could capture the essence of the big-data issue in such a sticky (and humorous way).

In my opinion, here are the 2 major challenges an organization usually faces.
1. Limited resources leads to a vital question of prioritization and the short-term ROI of analytics
Nowadays, many enterprises are still struggling to get their traditional reporting and BI systems to work effectively due to many old problems such as: lack of good data and silos data sources and analytics resources across different departments. While big organizations may have the luxury of dedicated R&D budget for experiment, small organizations will often face questions such as: should I invest $800K in SAS now, how much business impact it could immediately deliver and guarantee, and how could I measure? Especially, in the industries or the markets that either firms are still struggling with core operation or system issues or the culture has been traditionally no data-driven, to push a business case for analytics forward needs a roadmap of quick wins that could translate into $$, and therefore a very clear goal of what we want to do with insights is very important – at least in the short run until we could use the success stories of the quick wins to be the change agent to improve confidence of the organization and shift the culture more susceptible to innovation.

2. Culture & Talents
A research of Harvard Business Review showed that only 38% of 5,000 employees at 22 global companies were best equipped to make good decisions based on good interpretation of data. I believe that number is far less in Asia. It means that even if the organization has lots of data flying around, it does not bring any value as not many employees could understand and (believe on values of data) and act on it. To shift the organization to the next level, it will need talents and management commitment to cultivate data-driven culture with strong support of systems and methodologies to close the gap between insights (on paper) and execution.

Alicia Montoya - 24 Jan 2015, 1:05 p.m.

One year later, the discussion rages on. In case you missed it, take an hour to watch this WEF 2015 session on "The New Digital Context". I love YGL Max Levchin's view of data as a currency and (interconnected) devices as on-the-ground sensors, and his advice for us all of us: Disrupt yourself, constantly!

If you would like to leave a comment, please, log in.