Archive for May, 2018

Testing At the Speed of Now

May 28th, 2018

We read about artificial intelligence (AI) revolutionizing marketing, but most marketers are still asking “What is AI? And what exactly is it doing for marketing?”

One very practical answer is that AI has accelerated market testing. Obsolete are the elegant, three-month experimental designs — the pre-test, post-test, test-market versus control group. Today, AI enables marketers to test in real time and take advantage of the rapidly changing market/channel environment.

Control Groups in the Cloud

Much of the rigor of market testing comes from defining comparable test and control groups.

Back when TV advertising was the main marketing weapon, control and test markets were geographically defined — areas of dominant influence (ADIs), where TV advertising could be managed. Today, thanks to data mining and computer algorithms, test and control groups can be very accurately defined and maintained in cloud-based data sets.

Control and test groups can be created in which for every member of one group, a mirror image, twin customer is found for the other.

Comparability is also much more refined today. Computers armed with AI capture the hundreds of characteristics and behaviours of each customer as they occur moment by moment and segment customers by finding the patterns of demographic and behavioural characteristics. Computer learning algorithms, another modality of AI, continually sift and sort and analyze the roiling sea of digital transactions.

Short and Long-Term Control Groups

The experimental design process starts with identifying control groups for each in a series of marketing campaigns.

The next step is to define stable, long-term control groups. These control groups allow for multiple touches across multiple channels. They are larger in size than campaign-level control groups, which means they permit more robust and definitive conclusions than are possible at the campaign level. A stable, long-term control group provides the marketer with the ability to understand the impact of marketing across a series of campaigns over time.

Once the campaign-level and stable control groups are defined they need to be updated and maintained. Customers are constantly sending new signals. The AI algorithms detect and analyze them and enable the marketer to respond appropriately. Machine learning creates and constantly updates predictive models to gauge the likelihood of the customer’s next action. This modelling helps the marketer understand where customers are in their buyer’s journey and which offers are most likely to get them to purchase. The models enable the marketer to treat high near-term purchase probability customers very differently from those who are less likely to purchase in the near term.

With model-based testing, the marketer can set up and continually refine complex business rules where customer behaviour is interpreted as they respond to email, interact with digital properties or engage with social channels.

Once the models and rules are tested and established, the marketer can create data-driven conversations with customers. Their marketing strategies and campaigns can cater to individual customers in the moment in ways the marketer knows they will respond favourably to.

AI Is Just a Tool

While it is tempting to believe that AI enabled marketing can take the place of the marketing department, in reality, AI is just a tool. It needs to be used by experienced marketers with good judgment who:

  • Develop communications and campaigns that deliver real value. The marketer must understand who her customer is, so offerings can be informative, helpful, approachable, generous and, most of all, respectful of the customer’s precious time.
  • Have the confidence to move fast. Get ahead of the audience and competition with a strong bench of kits with trigger-based solutions — to deploy in real time based on test results.
  • Know when to update and evolve market segments based on AI and computer learning.

AI-enabled tools give the marketer the ability to:

  • Quantify the probability that specific customer behaviour actually precedes, leads to or signals a future behaviour.
  • Quantify uplift in desired customer behaviours and incremental impact on revenue generation.
  • Move fast to design, implement, assess and revise tests of her marketing strategy.

Rather than replacing the marketer, AI enabled testing at the speed of now helps marketers apply their expertise and experience to implement real-time marketing.


Source: All the above opinions are personal perspective on the basis of information provided by Forbes and contributor KN Kasibhatla.



Why You Must Treat Artificial Intelligence (AI) As A Very Special Technology

May 21st, 2018

There are lots of technologies that attract our attention – and money – these days. We’re obsessed with blockchain, cryptocurrency, IOT, big data analytics, cybersecurity3-D printing and drones. We’re excited about virtual reality, augmented reality and mixed reality. We love talking about driverless cars, ships and planes. We can’t wait for 5G and Wi-Fi domes that solve all of our network access problems; and while we’re getting a little worried about social media and privacy, we’re still addicted to our ever-more-powerful smartphones. We buy everything online. We’re into wearables. But there’s one technology that we all need to embrace: artificial intelligence (AI). While there are other families in the disruptive digital technology world, this one is special and one you cannot afford to treat as just another emerging technology. AI powers, amplifies and therefore supersedes them all.
Why so special?
First, AI is special because it’s more than one technology. In fact, it’s a family of technologies. Secondly, AI is special because its application potential is so wide. Next, AI is special because it learns and sometimes even self-replicates. AI’s also special because it satisfies ROI models of all shapes and sizes. Finally, AI is everywhere: which companies – and countries – are not investing in AI? There’s a bona fide arms race underway among the players (which shows no signs of slowing anytime soon).
What is AI?
AI includes at least machine learning, deep learning, image recognition, robotic process automation, natural language processing, text mining, vision systems, speech systems, neural networks and pattern recognition, among other methods, tools and techniques that according to the father of AI, John McCarthy, represent “the science and engineering of making intelligent machines, especially intelligent computer programs.”
What Can AI Do?
There is very little AI cannot do. The range of applications is staggering, including all of the vertical industries and every business process and model that supports them. AI will profoundly impact healthcare, transportation, accounting, finance, manufacturing, customer service, aviation, education, sales, marketing, law, entertainment, media, security, negotiation, war and peace. No industry or process is safe from the impact that AI – across all of its components – will have in the short-run and especially over the next seven to ten years. Keep in mind also that AI will integrate across business and technology architectures, data bases and applications.
What will AI change?
Everything. The timing – as always with the adoption of emerging technologies – is debatable. But the changes will not all be good. AI empowers good and evil. Note the ease with which fake news can be created and disseminated by intelligent “news” creators, and how easy it is for smart bots to service personal and professional confirmation biases intended to manipulate thinking and behaviour. At the same time, good bots will make much of our personal and professional lives more efficient and productive, freeing us to pursue other activities. Will AI eliminate jobs? Of course, and this time the elimination of jobs will include so-called knowledge workers as well as the traditional manufacturing jobs we associate with automation and robotics which will increasingly behave in unsupervised contexts. Much of this capability will arrive simultaneously across whole industries, such as the automotive industry which will utilize robotic AI to manufacture driverless cars and then manage their movement across cities and towns across the world. Similarly, healthcare will be impacted from lifestyles, monitoring, diagnosis and treatment. No, AI will not kill us, but it will augment and replace many of us in the workplace. Again, it’s a question of when, not if, but impact will be sweeping and will likely happen much faster than many analysts predict. Regardless of how bullish or bearish you are about displacement, it’s safe to say that tens of millions of jobs – and knowledge-based careers– will be impacted — and in many cases eliminated — in the next five-to-seven years.
Who’s playing?
Who isn’t? Investments in all things intelligent are unprecedented. All of the major technology companies are heavily invested in the technology, but the most important investment portfolio belongs to whole countries which have declared AI as a strategic national objective. China, for example, has defined AI as one of its core industries.
What to do about AI?
If your company is not already investing in AI, it’s way past time. Step one is the modelling of your current and aspirational processes informed generously by the potential of AI and predictions about the evolution of your industry. Elaborate process models should be developed, tested, simulated and inventoried to inform your AI pilot agenda. The simplest way to build this agenda is to identify the processes most amenable to AI and simulate the impact intelligent systems might have on the costs and benefits of the target processes. The most robust simulations should rank-order the processes that should be piloted with new technologies. Corporate partnerships should also be aggressively pursued, especially since AI is so broad. Companies need enabling partners that, for example, provide AI development and application platforms (which will come from their cloud providers in most cases). University partnerships are also valuable. Companies should befriend AI start-ups. Many AI technology companies will scan the start-up terrain for acquisition targets. Just as many established companies will scan the same environment for the same reason.

National governments should strategically commit to AI. This means that the national research laboratories – like the National Science Foundation in the US – should receive additional, directed funds to pursue a broad program of research and development that assures a global presence in the development and application of AI, which should be declared a 21st century moonshot.

Do you have the right AI talent in your company? If you administered an AI IQ test back at the ranch, how well would the team do? If your company is like most, you will need to invest in AI education and training starting with executive education about the strategic role of AI in your industry.

Finally, brutal process assessments are necessary to optimize AI. No process should be exempt from what AI might offer. But make no mistake, much of this is political. There will be Luddites who challenge the applicability and power of AI. But AI is different from the other “stand-alone” emerging technologies. AI can disrupt your business in ways you need to identify – before you’re disrupted. So do you need an AI “czar”? You absolutely do.

Source: All the above opinions are personal perspective on the basis of information provided by Forbes and contributor Steve Andriole.

Embracing AI and Automation Can Make Your Job Better

May 14th, 2018

By now, many people have heard of the impending “fourth industrial revolution,” and there’s more than a little trepidation surrounding the subject. Just as mechanization and the steam engine changed the landscape of manufacturing, the arrival of interconnected machine learning systems will inevitably transform the way products are made and sold.

The fourth industrial revolution may spark the fear that jobs will disappear. Emerging technologies will have a far-reaching impact, affecting almost every industry and economy on our globalized planet. However, artificial intelligence will serve in large part to augment – not replace – the jobs humans perform in the workplace. The emphasis on AI is to eliminate mindless busywork, making people more efficient, productive, and valuable.

John Carney, senior vice president of industries, communications and media at Salesforce, acknowledges there will be significant changes in the job market but explains that this has been the case throughout history: “If you look back in history with these paradigm shifts, these transitions, the data says that we created way more jobs than were eliminated. So, I think that’s going to happen again. There is going to be a transition.”

Forward-thinking employees should embrace this transition. Instead of being fearful of change, look to the following scenarios to consider how AI can improve your role.

  1. Marketers can finally give customers the experience they crave.

Unless you’re running a 24/7 call centre, your employees have to go home at some point. Enter AI, and the chatbot, which can give any number of customers the information they’re looking for at any time – even when the customer isn’t sure how to contact you directly.

Naveen Rajdev, chief marketing officer for Wipro Limited, a leading global information technology company, describes AI as the future of digital marketing, allowing business leaders to predict customer needs and respond. He sets the scenario: “It’s the middle of the night, and a customer mentions your brand in a tweet to ask about your store’s holiday hours. All of your employees are home asleep, and the customer assumes he’ll have to wait until morning for an answer. But then, just seconds later, he receives a personalized response answering his question.”

While AI won’t be able to answer every question, it will drastically reduce the burden on your human customer service reps. As a result, you can cut those lengthy hold times and make a big improvement to the overall customer experience.

  1. Programmers can focus more on the big picture and less on small jobs.

Programming can be tedious work, but AI is liberating programmers and allowing them to focus on the work that they enjoy. Praful Krishna, CEO of Coseer, describes how AI has improved his team’s workflows: “Our team members actually look forward to running modules that involve such tedious programming because now they can play with the AI, train it, monitor its results, and give feedback. It’s almost as if they have a team working for each of them.”

Utilizing AI has been a distinct win, but Krishna points out that there is a “trust curve.” New employees will manually check AI performance, which largely defeats the purpose of having it. Still, they eventually learn to trust it, and he reports that the process pays off in a matter of a few months or even less.

  1. Salespeople can shorten the sales cycle and bring in more qualified leads.

Sales departments have come a long way from simply making as many cold calls as possible. Today, there are many tools to ensure that valuable time is spent only on the most promising leads. AI has the potential to make this screening process even more precise and to impact the customer experience, improving the rate at which conversions are made.

AI can also take on time-consuming manual tasks that are nonetheless necessary. Jen Tadin of Gallagher, an insurance brokerage, explains how AI frees up time and improves productivity: “AI is used to prefill cumbersome underwriting questions that are required to get a quote. Much of the underwriting information is public domain; therefore, AI is critical in improving ease of doing business for both the prospect and company and assists us in closing more deals in less time.”

Using AI not only improves the experience and increases productivity for Gallagher employees, but it makes things easier and more convenient for customers as well.

  1. HR can speed up recruitment and training

Especially in large organizations, hiring hundreds or thousands of employees can create a massive backlog for HR. Combing through cover letters and résumés takes a huge amount of time that could be better spent elsewhere. Fortunately, AI can take on some of the burden.

Chatbots are able to perform basic background checks to immediately weed out undesirable candidates, while other AI tools can look for specific qualifications. Argentinian credit firm uses deep learning to screen applicants, saving HR employees as much as two-thirds of their time and allowing them to focus on higher-level tasks.

Cristian Rennella, the company’s HR director and co-founder, says he’s seen productivity increase by 21.3 percent: “We are sincerely surprised with the results and very happy because boring, manual, and repetitive tasks today can be performed automatically thanks to AI. And we can invest more time to interview the best candidates.”

Looking for more reassurance? Despite the proliferation of AI, the U.S. is continuing to experience record-low unemployment rates. The high-touch jobs that can only be performed by humans aren’t going away; humans still have a corner on emotional intelligence.


Source: All the above opinions are personal perspective on the basis of information provided by Forbes and contributor William Arruda.

Blockchain As An Application Platform

May 8th, 2018

Many business use cases can be improved and/or solved by using distributed ledger technology. It can be used in many cases where trust services are needed by business applications. This can be utilized by using blockchain technology as an application platform to build the underlying trust infrastructure of the system.

Although Bitcoin, the first real implementation of blockchain, is a decentralized currency and payment system, the underlying constructs that form the basis of the system do not have to be limited to payment transactions, accounts, balances or users. Instead, blockchain technology in Bitcoin is nothing more than transactions secured and executed by a scripting language using cryptographic methods. This means that blockchain is a platform with a scripting language that can solve many use cases other than just cryptocurrencies.

This property of blockchain led to smart contracts, an innovation presented by the cryptocurrency known as Ethereum. In the case of Ethereum, developers can create private cryptocurrencies and contract-based applications using a Turing-complete language, which allows businesses to use this language to set their own rules and policies in such applications.

The distributed ledger technology used in blockchain offers multiple benefits to businesses that make a difference when implementing a solution that requires a high degree of trust for business transactions. Using the technology offers the possibility to reduce costs and offers the opportunity for businesses to build and maintain an infrastructure that delivers capabilities at lower expenses than traditional centralized models.

Blockchain can process transactions faster because it doesn’t use a centralized infrastructure. Although there is no system totally secure from cyber attacks, the distributed nature of blockchain provides an unprecedented level of trust. The unchangeable property of blockchain and its public availability among its users, whether in a public ledger or a private one, provides transparency. Any user of the system can query transactions on a real-time basis.

Bitcoin was the first implementation of a cryptocurrency based on distributed ledger technology. It was invented in 2009. and since then, it has been gaining popularity and traction by business owners seeking a distributed trust model. The Bitcoin consensus algorithm is based on proof of work (PoW). In PoW, transactions are collected into blocks by miners and added to the blockchain only if the miner can solve a cryptographic challenge that requires much computational power to be solved. The cryptographic challenge can only be solved by guessing, ensuring neutrality.

Other forms of proofs have been invented and incorporated into other solutions, such as the proof of stake in Ethereum and proof of elapsed time introduced by Intel.

Bitcoin and blockchain solved a very old digital currency problem that many other digital currencies tried to solve in the past known as the double spending problem. Double spending means spending the same digital currency twice, and Bitcoin solved this by ensuring distributed consensus.

Another cryptocurrency benefit that blockchain technology provides is that transfers can cross national boundaries in seconds, with minimum fees, and without going through third-party entities such as banks.

The U.S. government and Venezuela are currently investing in resources dedicated to research and to create their own cryptocurrencies tailored to their specific needs. Despite the vast success of Bitcoin and other altcoins, the shortcomings in the design have limited the global adoption and expansion of cryptocurrencies. The expansion of cryptocurrency use will require overcoming governmental requirements and concerns, such as protecting against money laundering, illicit trades, volatile value and the lack of recognition by trusted parties.

Blockchain For Digital Identity

The need for a single centralized source of truth about identities is becoming a necessity in every community and corporation. Imagine a decentralized digital identity system, a source of truth where every single data element, such as user attributes and credentials, are included in the system only by distributed consensus.

This model is the focus of many enterprises, including Microsoft and IBM. Users get more control over their identity as they can share it only with trusted parties. No single centralized entity can tamper with user identities or data.

For users, this model improves accessibility, privacy of their data and control over their personal data. For enterprises, this model reduces identity management cost, eases the monitoring process, and improves customer service and efficiency.

Blockchain For Real Estate

Smart contracts in blockchain are little programs that execute if certain criteria are met. Smart contracts were invented in the 90s by Nick Szabo. They were integrated into blockchain technology and cryptocurrencies by Ethereum. In a smart contract, parties can agree on a sequence of conditional execution paths based on events. This idea led to the use of blockchain within industries such as real estate. Actually, smart contracts can work for any system that involves a contract between a seller and a buyer.

In the real estate industry, dealing with properties involves several parties and individuals, including owners, lenders, investors and service providers. The transactions between these entities can be problematic with the existing traditional centralized systems. This difficulty comes from many factors, including a lack of trust among peers, fraud and deficiency of a single source of truth about real estate and its history. Blockchain technology offers the possibility to have a real estate system with a very efficient search engine and lookup source for the current properties on sale.


As you can see, blockchain technology offers plenty of opportunities for various applications. And as the technology continues to progress, its applicability will only continue to broaden.


Source: All the above opinions are personal perspective on the basis of information provided by Forbes and contributor Forbes Technology Council.