Super to become ‘hyper-personalised’: Bravura

From Investor Daily.

Superannuation funds are in the process of collecting “unprecedented” amounts of member data that will be used to create hyper-personalised services, says Bravura.

In a new report titled Super Megatrends, Bravura superannuation product manager Scott Kendall said big data and predictive analytics are becoming essential tools for super funds that are looking to retain members.

A better understanding of member data can also improve the risk management and compliance processes of the big superannuation funds, Mr Kendall said.

“Modern technology platforms are facilitating the collection of unprecedented amounts of member data, enabling detailed member profiling in an effort to deliver hyper-personalised service offerings,” he said.

Super funds are employing techniques such as behaviour modelling to identify members who are at risk of leaving the fund, Mr Kendall said – and scenario stress testing can help funds better manager investment risk.

“As their use of big data becomes more sophisticated, all kinds of businesses are drawing upon information from other external data sources, such as social media and data aggregators, to gain an even more intimate knowledge of their customers,” he said.

Mr Kendall pointed to Mercer’s Harmonise platform in the UK, which provides an aggregated view of members’ finances through their employer.

“Participating funds have access to far greater amounts of member information than ever before and the additional data captured via this loop can be applied to deliver more meaningful, hyper-personalised services to their members,” he said.

Super funds will also start investigating artificial intelligence services such as ‘cognitive computing’, Mr Kendall predicted.

“Various insurers are already employing the services of IBM Watson to process large amounts of data to achieve better underwriting, fraud detection and credit control,” he said.

“It’s only a matter of time before super funds follow suit.”

Harnessing machine learning in payments

Good article from McKinsey on the revolution catalysed by the combination of machine learning and new payment systems as part of big data. The outline some of the opportunities to expand the use of machine learning in payments range from using Web-sourced data to more accurately predict borrower delinquency to using virtual assistants to improve customer service.

Machine learning is one of many tools in the advanced analytics toolbox, one with a long history in the worlds of academia and supercomputing. Recent developments, however, are opening the doors to its broad-scale applicability. Companies, institutions, and governments now capture vast amounts of data as consumer interactions and transactions increasingly go digital. At the same time, high-performance computing is becoming more affordable and widely accessible. Together, these factors are having a powerful impact on workforce automation. McKinsey Global Institute estimates that by 2030 47 percent of the US workforce will be automated.

Payments providers are already familiar with machine learning, primarily as it pertains to credit card transaction monitoring, where learning algorithms play important roles in near real-time authorization of transactions. Given today’s rapid growth of data capture and affordable high-performance computing, McKinsey sees many near- and long-term opportunities to expand the use of machine learning in payments. These include everything from using Web-sourced data to more accurately predict borrower delinquency to using virtual assistants to improve customer service performance.

Machine learning: Major opportunities in payments

Rapid growth in the availability of big data and advanced analytics, including machine learning, will have a significant impact on virtually every part of the economy, including financial services (exhibit). Machine learning can be especially effective in cases involving large dynamic data sets, such as those that track consumer behavior. When behaviors change, it can detect subtle shifts in the underlying data, and then revise algorithms accordingly. Machine learning can even identify data anomalies and treat them as directed, thereby significantly improving predictability. These unique capabilities make it relevant for a broad range of payments applications.

What is machine learning?

Machine learning is the area of computer science that uses large-scale data analytics to create dynamic, predictive computer models. Powerful computers are programmed to analyze massive data sets in an attempt to identify certain patterns, and then use those patterns to create predictive algorithms (exhibit). Machine learning programs can also be designed to dynamically update predictive models whenever changes occur in the underlying data sources. Because machine learning can extract information from exceptionally large data sets, recognize both anomalies and patterns within them, and adjust to changes in the source data, its predictive power is superior to that of classical methods.

The future of online advertising is big data and algorithms

From The Conversation.

The challenge facing advertisers and advertising professionals is remaining relevant in the face of a fundamental technological change. Namely, algorithms and big data.

The combination of the two, in the form of automated and real-time buying and selling, is redefining the advertising business model and value proposition.

Advertising is now a world of software, ultra-high-speed networks and processing power, statistics, optimisation, operations research, heuristics, data science and a range of related disciplines all coming together in dealing with large volumes of rapidly changing data.

How advertisers adapt will define their viability in the new world of online advertising. Period.

The trends marrying data and advertising

A number of interconnected trends are coming together to make this world of data.

Over 40% of the world’s population now has access to the internet. This is both a large market for advertisers to go after and a huge source of data. But the explosive uptake of smartphones has also brought on a large number of first-time internet users – fresh eyes for advertisers – and there are many more to come.

Global technology platforms, such as Google and Amazon, Facebook and LinkedIn, have created huge pools of data and made it all useful. Simply put, big data is data that’s too large or complex to be effectively handled by standard database technologies currently found in most organisations. But these platforms, among others, enable the data to be analysed and useful information extracted.

Historic data is also a potential gold mine in the right hands. It offers insights into industry trends and buying behaviours over time. Correlating historical data with “new data” can lead to the development of predictive models, also useful for advertising.

It’s not just the data we willingly give over to Facebook and LinkedIn that is useful, but behavioural, demographic, geospatial and other metadata as well. We are all leaving digital footprints each time we interact with the internet.

The global volume, velocity and variety of this data is astounding. This data, and the real-time insights and patterns that can be extracted from it, is the basis for tomorrow’s digital alchemy.

There is also a growing daisy-chain of intermediary organisations that harvest, analyse, interpret and offer up precisely targeted advertising services, all in near real time. At every stage, advertisers and intermediaries are clipping the ticket as they take their cut for being involved in the overall management of the torrents of data generated by us and fed back to us in the form of targeted advertising.

Putting it all together

Finally, there’s the rise of programmatic advertising – the real-time and automated buying and selling of ads with algorithms, bringing together many of these trends. There are now huge online marketplaces where software buys and sells advertising space.

Many of the ads you encounter around the web are now programmatic, allowing advertisers to target who sees an ad, based on this increasing array of data. This allows advertisers to predefine criteria for their ads – only showing them to Australians at a specific point in time, for instance.

As the internet transitions from an “open” and ostensibly free network to one that is ubiquitous and highly monetised, advertising is being catapulted into a new paradigm. The sheer value and growth of the online advertising market is reshaping the entire advertising industry, and the rate of change is not slowing.

The fundamental concepts of advertising remain unchanged. That is, to present the concepts, products and services of the advertiser that connects sellers with potential buyers.

What has changed, however, is that the advertisers (and sellers) must be able to harness the arsenal of real-time measurement and placement tools to focus their efforts with pinpoint accuracy and minimal cost. Being able to make use of big data, analytics and algorithms isn’t just “nice to have”, it’s essential.

Welcome to a new world of online advertising.

Author: Rob Livingstone, Fellow of the Faculty of Engineering and Information Technology, University of Technology Sydney

Banks and The Value of Their Customer Data

Interesting article in the SMH, discussing the leverage banks hope to get from the data they hold on customers. The explosion in digital banking has meant the banks’ information about their customers has ballooned. And advances in technology mean they can analyse it in ways that were previously impossible, to ensure they remain relevant.

Here is a snip-it:

… as banks eye the huge potential to enmesh themselves more deeply in consumers’ lives, and fight off lower-cost competitors.

Thanks to advances in computing power and customers’ embrace of digital finance, banks know more than ever about what their customers are up to: whether it’s browsing the web, shopping online, visiting the mall, or interacting on social media.

Already, they are busily harnessing this vast amount of data to sell products to customers before they ask for them: pushing travel insurance to someone who’s just bought airline tickets, or suggesting a home loan to the newlywed couple. But over the coming years, it is set to get much more tailored to the individual, and far more widespread.

As the traditional business of banking faces growing competition from new digital rivals, experts predict banks will increasingly be pushed into targeting customer “experiences” as they seek to remain relevant, and highly profitable.

Inevitably, however, this will involve a tension between what customers regard as the bank being helpful, and when it veers into the territory of ‘Big Brother’.

Follow the link to read the full article.

China’s plan to put two-faced citizens on credit blacklist isn’t all that foreign

From The Conversation.

China has a problem.

No, not Donald Trump trying to savage it any time he comes within three feet of a microphone. It’s that enormous social shifts in recent years – like the forcible relocation of 250 million people from rural areas to urban environments – have transformed the country, in the words of its Academy of Social Sciences, from “a society of acquaintances into a society of strangers.”

And these strangers, it turns out, don’t think much of each other. Social trust is at miserable levels, leading to a shaky business environment in which half of all written contracts are blatantly breached.

Since part of the problem is the lack of a credit reporting system, the government has decided to establish one. But instead of only considering people’s ability to repay loans, this system will rank people based on their trustworthiness using all sorts of data.

This might sound exactly like the kind of thing you’d expect from an authoritarian regime. And as someone who has pondered the ways in which privacy is squeezed by an ever-expanding surveillance state, I was intrigued by this unholy alliance between Big Data and Big Brother.

But what really surprised me was not just the outlandish lengths to which the Chinese government will go to evaluate its citizens. It was that its tactics were surprisingly close to what is already happening here, as banks look for ways to lend money to – and collect fees from – people with no traditional credit history.

But first let’s look at what the Chinese are doing.

The glories of trust-keeping

Using an enormous range of information, from traffic violations to consumer patterns to social networks, China intends to give every one of its 1.3 billion citizens a “social credit” score by 2020.

A recently translated summary of the plan explains that the goal is nothing less than raising “the sincerity and quality of the entire nation.” That, it says, should help address everything from workplace accidents to food safety failures to tax evasion and production of counterfeit goods (putting Canal Street, every New York woman’s go-to source for knockoff Chanel handbags, rather under a cloud).

The plan includes recommendations for establishing “civil servant sincerity dossiers,” something I’d like to see applied to my local DMV, lots of talk about “professional ethics, household virtue and individual morality” and encouraging companies to conduct “client sincerity evaluations.”

I’m not sure what that means, but it conjures visions of online retailers diligently making entries like, “Disappointing customer. Returned item saying ‘It didn’t fit.’ Strongly suspect she’s lying about being a size 6.”

There’s also a large public relations component, with the use of news media to “forge a public opinion that trust-keeping is glorious” and a raft of proposed holidays, including “Sincere Trading Propaganda Week” and “Quality Month.”

Alibaba’s Jack Ma wants to read your mind. Reuters

The pains of trust-breaking

Before you start worrying about the caliber of the other 11 months of the year, you’ll be glad to hear that there’s also a strategy for enforcement. This includes informants, blacklists and the rather chilling promise that “those breaking trust will meet with difficulty at every step.”

Interestingly, the government is letting private companies, like Alibaba, the e-commerce giant that made US$1 billion in eight minutes the other day, take the lead in a series of pilot projects.

Alibaba’s finance arm, Sesame Credit, has been issuing customers with social credit scores based in part on their purchases and hobbies.

As Sesame’s technology director explained, someone who played hours of video games “would be considered an idle person,” so less creditworthy, while someone “who frequently buys diapers” is probably a parent, so “more likely to have a sense of responsibility.”

Suddenly that puts Nicolas Cage in Raising Arizona, running from the cops with a stocking mask over his head and a package of Huggies under his arm, in a whole new light.

Raising China

Rank your friends!

Although it seems that someone’s score, rather shockingly, may rise and fall with the creditworthiness of their friends and relations, companies are focusing consumers on the positive.

Sesame has even launched a mobile phone game in which users can guess whether they have higher or lower scores than their friends. What could be more fun than seeing whether your friends are – literally – worth hanging out with?

This may all seem crazy, in ways both scary and silly. But before we get too smug about how it would be unthinkable here, consider the recent news about credit agencies “exploring new ways of assessing consumers’ ability to handle loans,” right here in the United States.

These include scouring “phone and utility bills, change-of-address records and information drawn from DVD clubs and suppliers of rent-to-own furniture.” And that’s just the well-known companies like TransUnion and FICO.

Start-up credit agencies and banks, reports The Economist, go even further, “piecing together scores by analyzing applicants’ online social networks,” monitoring their Facebook messages and determining whether they are spending prudently.

(Here we pause as I put down my phone, from which I was just about to order a gravy separator from Williams-Sonoma, in case I needed to separate gravy sometime. Suddenly, it just didn’t seem – what’s the word? – prudent.)

The credit agencies say that they are responding to a demand by their customers – the banks, which are looking for new sources of revenue and hoping to find it in people who previously had no credit score.

In China, diapers apparently means trustworthy. She deserves a loan. Reuters
These guys would have a tougher time getting credit. Reuters

Building a better citizen

So while we’re not subjected to a government effort to “build a better citizen,” as the Chinese are, we’re not doing much to prevent the private sector from conducting not-entirely-dissimilar data-mining investigations into millions of people too young, too poor or too new to the country to have traditional credit scores.

Ever since Target started using data-mining to predict whether female customers were pregnant (which explains why I received a can of formula, seemingly out of the blue, right before I had my first child), scholars have warned us about the many ways the private sector can use predictive analytics to figure out who we are and what they can sell us.

But even if it’s good business, there’s something odd about collecting all these disparate pieces of information – traffic violations, bills paid and unpaid, staying friends with your ne’er-do-well elementary school classmate, having children, playing Call of Duty: Black Ops III – and assigning the whole mess a single numerical score.

Reducing all aspects of social and consumer life to a single unit of value seems to fundamentally misunderstand the complexity of human experience. Maybe remaining friends with a childhood buddy with a poor loan history does reflect on your own financial creditworthiness. But that friendship might also point to other things about you – your past, your loyalty or your willingness to help those in need – that cannot be assigned a numeric value along the same spectrum as whether you paid your gas bill.

Maybe an authoritarian single-party state can’t be that concerned with the dignity and autonomy (let alone the privacy) of its citizens. But at least the Chinese plan has been publicly circulated. Its “you will be trustworthy – or else” message might be a little alarming, but it’s not like it keeps you guessing.

We can’t really say the same for our own shadowy system of credit ratings. And if the market requires it, how long will it be before we all get evaluated based on whether our purchases are of the “responsible adult” or “idle slacker” kind?

Better start stocking up on the Huggies.

Author: Caren Morrison, Associate Professor of Law, Georgia State University