Connect with us

Big Data

Why data science must stand at the forefront of customer acquisition



Just how valuable can big data be for a business? Some analysts believe that simply increasing data accessibility by 10 percent can help the average Fortune 1000 company generate an additional $65 million in income.

Quality data can be even more valuable for new startups. When you have a limited marketing budget, you can’t afford to let your customer acquisition efforts go to waste — especially when in many industries, companies spend hundreds of dollars to acquire a single customer.

Unlocking the potential of data science and analytics will enable you to gain greater insights into your customers, allowing you to spend your marketing budget more efficiently. Here are some of the top reasons why data science should play a central role in your acquisition strategy.

Identifying signals of intent and creating predictive models

When it comes to making the most of your marketing budget, few insights are more valuable than discovering why a customer wants to buy a particular product or service. This intent is usually signaled through a wide range of resources, including Google searches, visiting shopping comparison sites or reading product reviews on your own website.

Big data helps identify when a particular user is engaging in these activities, indicating that they are more likely to become a paying customer with the right marketing push. Targeting the right person at the right time is usually a recipe for sales success. Over time, as data science determines the strongest signals of intent, your team will also be able to create predictive models that will allow you to consistently target those who are most likely to convert.

By focusing your customer acquisition efforts on the customers that are demonstrating signals of intent, you’ll be able to stretch your advertising dollars further and get a much greater return on investment. Analytics can even be used to predict future needs, allowing you to nudge current customers at the right time to encourage additional purchases.

Testing marketing strategies

Data doesn’t just help you identify those who are most likely to make a purchase; it can also help you fine-tune the strategies you use to guide potential customers through the buyer’s journey. A/B testing can be used to determine the effectiveness of everything involved in customer acquisition. From comparing email campaign copy to changing the location of your call-to-action button, these tests will enable your team to find ways to keep customers engaged until they make a purchase.

As an example of this, Google Analytics allows businesses to break down their e-commerce products based on product list performance. This data goes well beyond revealing how well a certain item sold. It can also reveal which items receive a lot of views, but few clicks, or even which items are most likely to be abandoned in a consumer shopping cart.

Identifying underperforming product lists can greatly improve your customer acquisition efforts. Data will help your team find the reasons why a particular product isn’t performing well, or help you decide to discontinue an unappealing product. In some cases, even something as simple as changing the display order can provide a boost in sales — and A/B testing will reveal the answers.

Improved segmentation yields improved targeting

Not all customers are created equal — while some may become lifelong devotees to your brand, others may only make a single purchase. A customer may not generate a profit from an initial purchase, but if their lifetime value outweighs the cost of acquisition, your company will be better poised for long-term success. Once again, proper use of data analytics can help you identify the right customers to target as part of your acquisition strategy.

The Lyric Opera of Chicago “used machine learning algorithms to take into account hundreds of dimensions at once” to better fine-tune their audience segmentation strategy. Even without predictive modeling, these algorithms examined a wide swath of data that described top opera-goers. Combining this data with lookalike modeling improved their conversion rate by 3.7 times with a high-value group.

Segmentation data will allow you to identify those individuals who deliver the highest average customer lifetime value, ensuring that those you reach with your customer acquisition efforts will deliver significant profits in the long run.

The potential of big data

Put simply, better data enables better decision-making. Leveraging the power of big data will do much more than help your marketing team create more effective advertising campaigns. It can help you adapt your web content, fine-tune your SEO efforts and even help you discover new potential customer groups, all by providing invaluable insights that you wouldn’t be able to discover on your own.

For companies that truly wish to maximize their potential for customer acquisition, it is clear that big data is the key to a successful future.

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Big Data

Shed light on your dark data before GDPR comes into force



Shed light on your dark data before GDPR comes into force

Also referred to as unstructured data, dark data is growing at a rate of 62% per year, according to IDG. By 2022, they say, 93% of all data will be unstructured.

Gartner defines dark data as, “the information assets organizations collect, process and store during regular business activities, but generally fail to use for other purposes”. Consisting of data from a huge variety of sources – emails, documents, instant messages, digital media posts, partly developed applications – or just information which isn’t being used or analyzed, its nomenclature makes it sound foreboding. With new regulations such as the GDPR coming into force, businesses must gain a clear understanding of the data they hold. For structured data, this is straightforward. But dark data is much harder to manage, stored across a distributed IT environment with no single owner.

A ‘bottomless lake of data’

Dark data tends to be text-based data, as well as video, audio files and images. It’s generated by a diversity of different sources, gathered from mobile devices, social platforms, apps and internal systems to name but a few.  Much of the data generated by the Industrial Internet and the Internet of Things is unstructured, so this also falls under the dark data shadow.

In the workplace, employees are responsible for generating a lot of dark data. In fact, says Sony Shetty from Gartner, “Across the enterprise, employees are blindly building a bottomless lake of data and, in many cases, a corporate mantra of ‘save everything, just in case’ is encouraging the behavior”. Think about the amount of data you, personally, generate, filter and store each working day – did you record your last conference call in case anyone missed it? Did you make it available as a podcast and save that, too? What about your customer calls – do you record them ‘for training purposes’ and store them as audio files? Do you have a chat function on your website and keep a record of the interactions, or use an instant message function on your desktop? One study found enterprises to be using almost 500business applications, each generating data.

All the data generated by this activity falls under the definition of dark data, and is stored across different devices, drives, desktops and SaaS platforms. Most of it will never see the light of day again. Employees leave – taking their passwords with them – customers move on, business priorities change, and no-one has the remit, the ability or the time, to remove the data.  The information quickly becomes out of date and inaccessible.

The need to understand data

Prior to the GDPR, dark data would have been an accepted part of legacy business. In the UK, the 1998 Data Protection Act didn’t provide any minimum or maximum period for data to be stored, so it would have been a case of ‘out of sight, out of mind’. Now, though, the GDPR requires businesses to gain an in-depth understanding of how data flows across their organization, along with stringent data governance. The new Data Protection Bill coming into force will implement the GDPR into UK law. From May 25th, if a ‘data subject’ – a client, employee or other stakeholder – asks what data a company holds on them, the company must know and share this. If they ask to see a record of when and how they gave their consent to be used, the company must provide this too, and only information necessary for its original purpose should be processed. “Inaccurate or outdated data should be deleted or amended and data controllers are required to take “every reasonable step” to comply with this principle”, says Debbie Heywood from Taylor Wessing.

This is extremely hard to fulfil if data is held in silos across an organization.  “Because unstructured data is text heavy and irregular, making sense of what is being said and how it’s being said — posi­tively or negatively — is not for the faint of heart,” says a report from the Medallia Institute.

Tapping into uncharted territory

The time has come for businesses to bring their dark data into the light. Doing so helps drive GDPR compliance, but the benefits of understanding dark data stretch far beyond compliance. Think of it as discovering uncharted territory: analyzing this unstructured data offers the opportunity to extract invaluable business insight which would otherwise lie dormant. It transforms information from data into strategic intelligence. Gartner cite, “Some examples of data that is often left dark include server log files that can give clues to website visitor behavior, customer call detail records that can indicate consumer sentiment and mobile geolocation data that can reveal traffic patterns to aid in business planning.”.

For example, most of us know that retailers are experts at using psychology to drive product placements. They understand our thought process and how we tend to move around a store, and place products accordingly. Studying filmed footage of consumers’ mobility in stores helps retailers refine their product placement strategies even further. As Deloitte says, “A retailer may be able to gain a more nuanced understanding of customer mood or intent by analyzing video images of shoppers’ posture, facial expressions, or gestures”.  This intelligence, extracted by analyzing dark data, can translate directly into revenue as retailers apply it to their store layout.

By analyzing dark data businesses can:

  • Create a truly 360-degree single customer view, to drive engagement and boost interactions
  • Anticipate, understand and respond to changes in market- and consumer-demand
  • Develop an in-depth understanding of consumer sentiment on their brands, gleaned from social platforms and multichannel interactions
  • Lockdown and secure vulnerable data points, and give personal data the protection it requires
  • Refine the accuracy of risk management models
  • Address recurring pain points for customers and direct customer support to those areas most affected
  • Identify any links and connections between data sets
  • Generate a strong foundation for accurate forecasting
  • Gain a deeper understanding of website performance from web analytics
  • Identify new revenue streams. According to IDC, “By the end of this year, according to IDC, “50% of Large Enterprises Will Be Generating Data-as-a-Service (Daas) Revenue from the Sale of Raw Data, Derived Metrics, Insights, and Recommendations”.

Now, analyzing unstructured dark data is simpler than ever before. Advanced, high-performance Customer Information Management tools automate and accelerate processes, connecting data sets for clarity and insight. Software scans both structured and unstructured data, using different data profiling techniques. The results of the scan are used to automatically generate a library of documentation, which describes a company’s assets and creates a metadata repository. You can then start to explore the opportunities and possibilities which lie within the data – and that’s when it starts to get really exciting.

Continue Reading

Big Data

17 best data science bootcamps for boosting your career



17 best data science bootcamps for boosting your career

Data scientist is the best job in America, according to a survey on Glassdoor, and it consistently makes the top of the list year after year. With a job score of 4.8 out of 5, a median base salary of $110,000 per year and over 4,500 current job openings, it’s a great time to be a data scientist.

But, as the role of data scientist grows in demand, traditional schools aren’t churning out qualified candidates fast enough to fill the open positions. There’s also no clear path for those who have been in the tech industry for years and want to take advantage of a lucrative job opportunity. Enter the bootcamp, a trend that has quickly grown in popularity to train workers for in-demand tech skills.

Here are 17 of the best data science bootcamps, designed to help you brush up on your data skills, with courses for anyone from beginners to experienced data scientists.

The 12 best established data science bootcamps

  • Byte Academy
  • DataCamp
  • The Data Incubator
  • The Data Science Dojo
  • Dataquest
  • Galvanize
  • General Assembly
  • Level
  • Metis
  • NYC Data Science Academy
  • Springboard
  • Thinkful

Byte Academy

Byte Academy offers full-time, part-time and remote programs in data science; and attendees have the option to defer payment until they secure a job after graduating. If you aren’t hired within six months, Byte Academy will refund your full tuition. The full-time course is five days per week for 14 weeks, while the part-time course takes place two evenings per week over 24 weeks. Byte Academy also offers corporate and onsite training for companies and custom corporate training with topics like blockchain and quant-algos.

Switchup rating 4.62
Cost $14,950 for full- and part-time bootcamps, $5,500 for a la carte options. If you aren’t hired within six months after graduation, your tuition will be refunded
Locations New York City, Bangalore and online


DataCamp is entirely online and it’s aimed at professionals who are already working in technology, finance and healthcare. However, anyone interested in data science will benefit from DataCamp’s program. The courses not only teach you the necessary skills, you can also practice and apply those skills to real-world problems through hands-on projects. It’s free to try, but for full access you’ll need to pay a subscription fee.

Switchup rating 4.22
Cost Free with limited access or $29 per month/$25 per year for an unlimited subscription
Locations Online

The Data Incubator

The Data Incubator is an eight-week program aimed at more experienced tech workers with a masters or Ph.D.; fellowships are available for qualifying students. Qualified fellows “already have the 90 percent difficult-to-learn skills” and Data Incubator promises to equip them “with the last 10 percent.” The program also offers students mentorship directly from hiring companies, including LinkedIn, Microsoft and The New York Times, all while they work on building a portfolio to showcase their skills.

Switchup rating 4.71
Cost Free for those accepted
Locations Boston, Washington (D.C.) and online

The Data Science Dojo

With campuses in Seattle, Silicon Valley, Barcelona, Toronto, Washington and Paris, the Data Science Dojo brings quick and affordable data science education to professionals around the world. It’s one of the shortest programs on this list, but in just five days, Data Science Dojo promises to train attendees on machine learning and predictive models as a service, and each student will complete a full IoT project and have the chance to enter a Kaggle competition.

Due to the short nature of the course, it’s tailored to those already in the industry who want to learn more about data science or brush up on the latest skills. However, it’s open to anyone at any skill level — if you’re ready to throw yourself in the trenches of data science.

Switchup rating 4.88
Cost $3,000
Locations Seattle, Silicon Valley, Washington (D.C.), Paris, Chicago, Toronto, New York City, Barcelona, Amsterdam, Austin and Singapore


Dataquest is the most highly rated data science boot camp on review site Switchup, earning near five-star reviews for overall experience, curriculum and job support. Dataquest relies on teaching students through “interactive coding challenges” instead of video lectures, so you will have the chance to code and work with data, while receiving feedback as you go. You can opt for the Data Scientist, Data Engineer or Data Analyst path — each promise to prepare you to jump right into a career in data.

Switchup rating 4.96
Cost Free with limited access, $29 per month for a basic account and $49 per year for a premium account
Locations Online


Galvanize offers a 13-week, full-time data science course that focuses on everything you need to know to become an effective data scientist. You’ll learn the fundamentals of data science, including python, statistics and machine learning through real-world case studies. The course ends with a capstone project and you’ll be more than prepared to enter the job market. The application process includes a technical exercise to gauge your current level and to see whether you have potential to succeed in the field.

Switchup rating 4.21
Cost $16,000 for the 13-week course
Locations Denver, San Francisco, Boulder, Seattle, Austin, Phoenix and New York City

General Assembly

General Assembly offers full-time and part-time courses, workshops and events in person and online. The full course catalog is extensive, and there is a program for every data science skill you can imagine. Courses range from one-week accelerated courses to full-time immersive 10- to 13-month programs, but it’s easy to find something to fit your schedule and budget. Whether you’re a recent graduate, looking to make a job switch or you’re an experienced data scientist trying to expand your skillset, General Assembly will have a program for you.

Switchup rating 4.18
Cost Payment varies depending on the program you choose, but financing options are available
Locations Boston, London, Los Angeles, New York City, San Francisco, Sydney, Washington (D.C.) and online


Offered through Northeastern University, Level is a two-month program that wants to turn you into a hirable data analyst. Each day of the course focuses on a different real-world problem that a business might face, and students develop projects to solve these issues. Students can expect to learn more about SQL, R, Excel, Tableau and PowerPoint and walk away with experience in preparing data, regression analysis, business intelligence, visualization and storytelling. You can choose between a full-time eight-week course that meets five days a week, eight hours a day and a hybrid 20-week program that meets online and in-person one night a week.

Switchup rating 4.41
Cost Varies by programming, location and the schedule you choose
Locations Boston, Charlotte, Seattle, San Jose, Toronto and online


Metis has campuses in New York and San Francisco, where students can attend intensive in-person data science workshops. Programs take 12 weeks to complete and include on-site instruction, career coaching and job placement support to help students make the best of their newly acquired skills. Like other boot camps, Metis’ programs are project-based and focus on real-world skills that graduates can take with them to a career in data science. Those who complete the program can expect to walk away with in-depth knowledge of modern big data tools, access to an extensive network of professionals in the industry and ongoing career support.

Switchup rating 4.88
Cost Cost varies depending on the program, starting at $2,350 for in-person professional development and $1,900 for Live Online professional development courses. Scholarships available
Locations Chicago, New York City, Seattle, San Francisco and online

NYC Data Science Academy

The NYC Data Science Academy is aimed at more experienced data scientists who have earned a masters or Ph.D. degree. Courses include training in R, Python, Hadoop, GitHub and SQL with a focus on real-world application. Participants will walk away with a portfolio of five projects to show to potential employers as well as a capstone project that spans the last two weeks of the course. The NYC Data Science Academy also helps students garner interest from recruiters and hiring managers through partnerships with businesses. In the last week of the course, students will participate in mock interviews and job search prep; many will also interview with hiring tech companies in the New York and Tri-State area.

Switchup rating 4.92
Cost $17,600
Locations NYC and online


The Data Science Career Track at Springboard consists of a six-month program that typically requires 10 to 15 hours per week of work. You’ll get access to the Springboard community, a personal mentor, career coach and student advisor. By the end of the program, participants will have an “interview-ready portfolio” and access to a data science network. Most students complete the course in two to four months and Springboard promises to refund your tuition if you don’t land a job within six months after graduating.

Switchup rating 4.89
Cost $7,500 total, with month-by-month payment options
Locations San Francisco and online


Thinkful offers a self-paced online bootcamp with a project-based curriculum, career prep, and one-on-one mentorship with access to a full community of students, mentors and alumni. The course requires about 20 to 30 hours per week of work and most students graduate in about six months. Thinkful also offers a job guarantee — if you can’t find a job after graduating, the company will refund the cost of your tuition.

Switchup rating 4.79
Cost $7,999 upfront, or $1,495 per month with the option for loans, financing and other payment plans
Locations Washington (D.C.), Portland, Dallas, Los Angeles, Phoenix, San Diego, Atlanta and online

5 emerging and niche data science bootcamps

  • Data Science for Social Good
  • Data Application Lab
  • Insight Data Science
  • K2 Data Science
  • Microsoft Research Data Science Summer School

These data science bootcamps are recommended by Switchup, but they haven’t received many — or, in some cases, any — reviews. But that doesn’t mean they aren’t worth considering. Some are newer programs that recently launched, while others involve exclusive fellowships or target niche markets, like Ph.D. students or experienced professionals. Look to see if any of these data science bootcamps better suit your professional needs.

Data Science for Social Good

This Chicago-based bootcamp has specific goals; it focuses on churning out data scientists who want to work in fields such as education, health and energy to help make a difference in the world. Data Science for Social Good offers a three-month long fellowship program offered through the University of Chicago, and it allows students to work closely with both professors and professionals in the industry. Attendees are put into small teams alongside full-time mentors who help them through the course of the fellowship to develop projects and solve problems facing specific industries. The program lasts 14 weeks and students complete 12 projects in partnership with nonprofits and government agencies to help tackle problems currently facing those industries.

Switchup rating No reviews
Cost Free for those accepted
Locations Chicago, with plans to expand to Charlotte (N.C.) and Chile by 2019

Data Application Lab

Data Application Lab is another in-person, full-time data science bootcamp stationed in Los Angeles and Silicon Valley, but they also offer online options if you can’t get to those locations. The programs focus on equipping students with “industry practical needs” in conjunction with traditional academics. Courses use lectures, hands-on experience and lab projects to help expedite the learning process. If you have less experience in computer science, they also offer a programming class and a 10-week internship to get you up to speed.

Data science programs include general data science, big data engineer, the basics, data analyst, data science full stack and big data solutions. Each bootcamp has different requirements, so you’ll want to make sure you meet them before you apply. Tuition varies depending on the program, location or course you choose.

Switchup rating No reviews
Cost Free
Locations Boston, Washington (D.C.) and online

Continue Reading

Big Data

Reality check: the complexities of mapping the world in 2018



as a critical component of purchasing precise and relevant third-party data. Across the thousands of data sets I’ve examined over my career, many have contained major gaps or inaccuracies not initially apparent on the surface. However, vetting these data sets is just one of the challenges organizations face today.

What if there is no data to even buy? What if no creditable information exists in the areas you need to know more about?

With the exponential rise of physical, digital, mobile, and transactional data, many believe that complete, up-to-date, and reliable data—about anyone, anything, or any place—is readily available. Well, I’m here to tell them that they’re wrong. This information is simply not as obtainable as they think.

Exploring the demand for data

When you look at the origin of the data being collected by today’s businesses, it its being generated by people, connected devices, and activities. It is captured and subsequently made useful because there is a viable business demand for that generated data. In turn, it is made more widely available down the road once that data is offered to buyers at a reasonable price.

When we examine those regions with the greatest amount of readily available data, there are often three common attributes. These data-dense areas have:

  1. A large population of both people and businesses
  2. Fewer government data regulations, and often government involvement in data creation and publishing
  3. Low data purchasing costs

Areas that lack one or more of these essential factors will understandably have less data to work with.

Comparing data collection around the globe

Let’s take for example the United States. The vast majority of U.S. states are well populated, host many industrialized and data-driven organizations, have few data regulations, and – because of the Freedom of Information Act – have plenty of government-created data that can be used as a starting point for commercial offers. This combination results in commercial data being offered at lower prices relative to the rest of the world. Because of this, there are vast amounts of data about the American population at large.

Compare this to rural Africa, which has very low, centralized populations, and lacks a formal and modern workforce. Today, little data—or shall I say, little reliable data—exists about Africa for many of the business applications that U.S.-centric data users have come to expect.

If we look at China, with the largest population in the world and one of the most sophisticated, modern workforces, it is assumed that there is an incredible amount of data, and a strong business demand for that data. However, China has some of the strictest data regulations in the world, making it illegal for organizations outside the country to access and export that data from China.

The U.K., while home to some of the world’s largest data-driven organizations, and some of the most up-to-date, complete, and visually beautiful data, charges a pretty penny for the Crown-copyrighted data, making it difficult for most to access if they have a U.S. price reference in mind.

And new regulations with GDPR are adding to access complexities, as many organizations are still coming to terms with what data can be shared, and in what capacities.

Expectations for data today

We’ve found ourselves at a pivotal time when it comes to data collection, especially as analytics and machine learning fuel more and more business decisions. While our expectations are that the whole world is mapped, counted and described to the same level, the reality is that it isn’t. Describing the world through data is subject to many factors, and with the introduction of GDPR and recent public data security breaches, people, companies, and businesses are becoming more conservative than ever before when it comes to sharing information.

While data has inarguably reshaped the way we visualize the world, our greatest vantage point is still ahead of us. As organizations begin to get into a rhythm of what access to data and compliance now looks like with GDPR, the ways in which we visualize data will certainly change. Along with that, consumers’ understanding of the new regulations will vary. It will take a greater level of comfort when it comes to sharing data before we ever have a completely holistic view of the world.

Continue Reading