4imprint, LLC

| Updated: May 19, 2021

How to make sense of unstructured data

Businesses collect more data about their customers, their activities, their competitors and their markets than ever before. A digital record exists of nearly every event, from making a cold call to a prospect, to replying to a customer service question on Twitter®.

So much of the data created can’t be shoehorned into categories in a customer relationship management system (CRM). These voices and opinions say so much, but too often the best we can do is slot them under “Comments” and scratch our heads over what to do with them.

Those unstructured data don’t have to go unexplored. The answer is text analytics:

“Text analytics turns unstructured text into categorized information that can be analyzed and more easily explored; it helps us make sense of amorphous data and leverage the insights it contains to become MVPs in the eyes of our customers.”[1]

There’s no one way to practice text analytics, and the insights mined from unstructured data are different for each unique organization. Text analytics can help tell the story of your data and reveal new opportunities for your organization, thanks to analysis of unstructured data.

By 2020: This market will reach $6.5 billion and increase at a rate of 25% annually.

Chart showing the text analytics market will increase 25% by 2020

What is text analytics?

From a high level, text analytics processes unstructured data, which are free-form text (among other examples to be covered). You can think of text analytics as having three defining characteristics:[2],[3],[4],[5]

  • Text analytics focuses on unstructured data (vs. category-driven structured data you’d find in a CRM).
  • Text analytics extracts relevant information from bulk unstructured data.
  • Text analytics organizes unstructured data to reduce time to insight.

Giving structure to unstructured data is the goal of text analytics. When unstructured data are organized, two things can happen:[6]

  1. Unstructured data can be combined with structured data in an existing database. The whole collection of data can be analyzed for predictive or discovery purposes.
  2. Unstructured data can stand on its own for analysis. Once in a manageable framework, the data can help determine relationships and trends.

Structured data, the sort of data that can be found in a CRM, can tell an organization what is happening. To compare, unstructured data can tell an organization why it’s happening.[7]

What’s the value of text analytics? The insight gained from text analytics can drive a customer-centric business paradigm and reinforce the best parts of your brand.[8] The unstructured data processed in text analytics often is as close as an organization can get to the honest-to-goodness truth about its brand experience.

The value that can be created, thanks to text analytics, is tough to oversell. Let’s dive deeper into how text analytics can be put to work for different types of organizations:[9]

  • Financial trading: Perform domain-specific sentiment analysis and classification. Take cues from traders on when and how much to trade.
  • Voice of the consumer: Scan social media. Address gripes before customers call to complain or cancel.
  • Manufacturing or warranty claims: Review text in warranty claims, dealer technician lines, report orders, customer service reports and more. Identify trends for repairs or damage to determine whether to launch a recall.
  • Lead generation: Scan social media for people interested in a certain product or service. Share with sales for follow-up.
  • Recruitment: Get to know potential hires before they come on the market. Analyze their social media postings for cultural fit and expertise.
  • Product or service review sites: Cut to the chase to help readers. Use text analytics to condense valuable customer reviews to two to three word phrases.

There are almost as many different uses for text analytics as there are types of data. Marketing teams can use text analytics for churn analysis and market research. A human resources department can use text analytics in discovering the voice of the employee. Depending on the industry, text analytics can be used for fraud detection or medical research.[10] Text analytics makes unburying hidden treasure easier.

Getting to know unstructured data

To understand text analytics, you have to understand unstructured data. Unstructured data are found in many places: [11],[12]

  • Survey verbatims
  • Inbound customer communications
  • Review sites
  • Facebook®, Twitter and all social media communication
  • Videos and video transcripts
  • Emails
  • Reports
  • Spreadsheets
  • Contracts
  • Warranties
  • Telephone/member listing books
  • Advertisements
  • Marketing materials
  • Annual reports
  • Customer call logs
  • Employee evaluations
  • Ordering information
  • Blog posts

Unstructured data are all data that don’t neatly fit into one category or multiple choice answers. We’re talking answers to open-ended questions and unsolicited responses. Unstructured data are insightful but challenging to analyze because of their inherent messiness.

Each source of unstructured data has unique strengths and weaknesses that can’t be ignored. For instance, social media comments are unconstrained and unprompted, but you usually hear the voice of the extremes—people who were delighted or furious enough to sound off. Review sites share candid customer experiences, but many comments are fake or the result of the company encouraging customers to leave positive comments. Customer surveys often are more representative, but free-form answers still are influenced by the wording of open-ended questions. Text analytics provides a framework for looking at unstructured data and mitigating such biases.[13] Providing structure to information gives a better picture of what’s really going on.

Most of the data generated and collected are unstructured data. The popular estimate is that unstructured data account for about 80 percent of all business information.[14],[15],[16] And, about 70 percent of unstructured data is generated by customers.[17] Pretty much all business activity creates data that need some preliminary processing before it can be analyzed.

The digital universe of data is exploding. By 2020, the digital universe will grow by a factor of 10—expanding from 4.4 trillion gigabytes in 2013 to 44 trillion. Every two years, this universe more than doubles.[18] The greatest challenge of the modern organization could become information management.

Even though the overall amount of data is growing, the amount of data that truly matters to a business is surprisingly manageable. Organizations are looking for “target rich” data—that which is easy to access, available in real time, has a big footprint, is transformative and possesses intersection synergy (i.e., it has more than one “target rich” characteristic). Based on these criteria, only about 1.5 percent of the digital universe as a whole is “target rich.”[19] Using the “target rich” criteria makes gathering data feel less like searching for a needle in a haystack.

One of the largest obstacles to text analytics is processing language. Analytics software, however, is advancing and getting better at drawing information out of text comments.[20] Humans still are the best at picking up language nuances, but machines are improving.

It’s no surprise that as the amount of data has grown, so has the market for text analytics products. The global market for text analytics is projected to reach $6.5 billion by 2020, with a 25 percent annual growth rate.[21] Solutions are rising to meet the data deluge.

Why unstructured data matter

There are three major reasons why unstructured data—and consequently text analytics—matter to organizations:[22], [23],[24],[25],[26],[27],[28]

  1. Unstructured data are human. Humans want to tell stories; they don’t want to be bound by rigid scales and finite answer options to describe an experience. Unstructured data are the tales we tell, rich with emotion and detail. They hold vital information about the customer journey and help organizations better respond to needs. In addition, unstructured data can shape personalized user experiences, an expectation of more and more customers. Text analytics unlocks insight into how to create a more human experience.
  2. Unstructured data are increasing. Sources for marketing data in today’s 24-7-365 business world include new channels, competition, segments, product lifecycles and greater price transparency. We have never captured more data, and it’s never been more accessible, thanks to the Internet. But the epic volume of unstructured data, if left untranslated into actionable insights, “is as useless as salt water to someone who is parched and adrift on the ocean.” Text analytics organizes unstructured data and extracts insight; the result is fewer undiscovered business opportunities left on the table.
  3. The best in class use unstructured data. Data-driven decisions drive profitability. Organizations in the top third of their industry that are practicing data-driven decision making are 5 percent more productive and 6 percent more profitable than competitors. Businesses that use diverse data sources, analytical tools and metrics “were five times more likely to exceed expectations for their projects than those who don’t.” By managing unstructured data, text analytics is a tool to drive business goals and stay competitive.

As the amount of unstructured data increases, so do the number of ways an organization can leverage insight to improve its products and services. Text analytics holds the key to making this transformation in a scalable and efficient way.

Types of text analytics

So what does it “look” like when text analytics are performed on unstructured data? Unstructured data are diverse, and organizations are searching for different flavors of insight. Here are a few different types of text analytics:This image explains the types of text analytics.

  • Sentiment analysis: Analyze the opinion or tone of text. The first level is polarity analysis (positive or negative). The second level is categorization (e.g., confused, angry). The third level is putting the emotion on a scale (0-10). The next evolution in sentiment analysis will be comparing sentiment of different time periods and analyzing the birth of a trend. [29] The tricky part of sentiment analysis is identifying sarcasm, irony and context of communication.
  • Topic modeling: Identify dominant themes in collection of text. One way to model is known as latent dirichlet allocation, “where words are automatically clustered into topics, with a mixture of topics in each document.” Another way to model is called probabilistic latent semantic indexing, “which models co-occurrence data using probability.” Be aware that topic modeling doesn’t scale well.
  • Term frequency-inverse document frequency (TF-IDF): Discover the relative frequency of a word in a set of documents. TF-IDF can be predictive. For instance, to improve retention, an organization could analyze call center transcripts of former customers. Note of caution: TF-IDF can negate word order and phrasing and words that are similar, like hotel and motel.
  • Named entity recognition (NER): Use surrounding text to recognize nouns, such as people, organizations, places and dates. Be aware that it takes significant data prep and training for NER to work.
  • Event extraction: Discover relationships between nouns, and go beyond the conversation. This type of text analytics can “assign roles to entities, assign subtypes and link to semantic data.” Like NER, preparation for event extraction is extensive.

The wonderful thing about text analytics is that the practice can process volumes of unstructured data faster than a human brain. The less-than-wonderful thing about text analytics, however, is that because a human brain isn’t doing the work, the practice misses many nuances.

Text analytics case studies

Companies have used insight from this type of data to grow and improve processes. These are just some organizations that have tapped the power of text analytics:

Internal improvement

Deloitte®, an international management and financial services firm, uses text analytics to improve document usability for employees.[30] Workers create collections of documents based on their interests by “liking” the documents. Workers also can add comments to documents, and comments become discoverable by other users who share the same interests.

“Depending on how the user interacts with the document, the system will get trained on what related topics are also of interest,” said Ben Johnson, senior manager at Deloitte. “In that respect, we are approaching a machine learning model.”[31] The more employees use the text-analytics-drive system, the more useful it becomes.

Combining structured and unstructured data

Climate Change Capital℠, a financial asset management firm, integrated the unstructured data in its content marketing with the structured data in its customer relationship management system.[32]

“[We] aggregate our website’s content to show us which of our topics are getting more interactions from visitors,” said Climate Change Capital’s Carlos Moran, who heads the marketing and communications activities. “By organizing the information we have about it and feeding it into our CRM system, it allows us to draw insight from it and act upon it.”[33]

In a similar fashion, The New York Times® uses natural language processing, a text analytics tool, to discover content topics that engage the most readers. The Times’ marketing team then knows which types of articles to promote. The newspaper, however, still relies on its editorial staff to develop and choose content.[34] It’s through text analytics that organizations can unlock previously undiscovered paths to creating more value.

Product development and marketing

A text analytics system was used to evaluate 7,000 customer reviews of Amazon’s® Kindle Fire® HD. Analysis revealed positive sentiments associated with the screen, speakers, Android® operating system and apps. However, users expressed negative sentiments associated with the failure to provide a charger with the device.[35] Amazon could use this insight to improve the next generation Kindle Fire HD. And, the e-retailer has a deeper understanding as to which features elicit responses from customers. This information is a goldmine for shaping marketing messages.

Building your text analytics toolbox

Choosing tools and vendors for text analytics depends on your business goals, budget and internal resources. No one-size-fits-all solution exists. But, these elemental questions should be asked when vetting a text analytics solution:

1. How does this solution integrate data across systems?
The combination of structured data and unstructured data provides richness to your business activities and customer interactions. Having all relevant information in one place lets you get to the “why” of your data faster.[36]

Once all the data are in the same place, the solution should allow your team to view the data at different levels—from the proverbial 30,000 feet to individual customers.[37] Across the organization you’ll be able to identify opportunities for systemic improvements and retaining customers.

Legacy analytics software can make connecting unstructured and structured data difficult. Aim to automate this union as much as possible. Tools such as RSS, semantics analytics and metadata can (respectively) aggregate, analyze and standardize.[38]

When data are fragmented in different departments and divisions, the experience for customers also can be fragmented.[39] Data that work together can be leveraged by your team for the sake of your customers.

2. How does this solution help tell the story of the data?
A story-driven reporting solution helps explain the insight from your text data.[40] Stories are relatable, emotional and memorable, which makes putting the insight from them into action less intimidating and more natural.

Keyword extraction, pattern analysis and entity extraction (identifying people, companies, cities, etc.) are more organic ways to sort through text and starting points for storytelling. They help identify areas of interest.[41],[42] These tactics can reveal themes in the story of the data.

Text analysis can tell scary stories: which customers are about to flee, how you’re noncompliant or possibly facing legal action, or which products/services have quality issues.[43] These stories could be tough to hear, but they’re actually opportunities for improvement.According to Tom Reamy, who heads the KAPS Group, which specializes in text analytics consulting and developing semantic infrastructures:[44]

“Things like pattern recognition and concept identification can be done very readily. What is lacking is a deeper analysis and understanding. The meaning inside the text and the context around words are critical to text analytics.”

Context and insight come alive through your data’s storytelling. Stories give meaning, and an organization can shape its offerings in response.

3. How does this solution drive personalization?
Personalization is critical to win and keep customers, but execution is a struggle. The main challenge is developing a single customer viewpoint. North American companies use an average of 36 different data-gathering systems and vendors. But, only 24 percent of senior execs said the systems were integrated across their organizations.[45] A single customer viewpoint ensures that your brand is consistent no matter how a customer is interacting with your organization.

Organizations create silos with divisions, departments and teams, yet customers only experience one brand. Taking data from different sources and using it to develop touchpoints creates a fragmented customer experience.”[46] Data systems that don’t connect can be used only to inform a macro view of consumer trends. On the other side, data systems that allow drilling down for granular insight make personalization easier.

Unstructured data can be used to understand customers and to understand every touchpoint in the customer lifecycle.[47] This builds a personalized experience that is shaped by who they are as individuals and what we deliver to them as an organization.

Integrating data sources can help firms develop a single customer view. All the facts about all types of experiences live in one place; it’s easier to make decisions with the full context.[48]

Ideal state: The full data picture

Companies have access to many sources of unstructured data, and have many ways to put them to work. Text analytics is the key to transforming data to insight. The right text analytics solution for your organization will aggregate, analyze and standardize all data and help tell the story of the data.

With the volume of unstructured data exploding, organizations have more opportunities than ever to discover new ways to personally delight customers. Yet, a business will struggle to deliver personalization if it doesn’t integrate structured and unstructured data.

The best decisions about your business will be made with the fullest context possible. Text analytics can help you get to the story behind the story of your data—and write your next chapter for success.

Endnotes

[1] Turner, Michelle. “Channeling Billy Beane.” Editorial. Marketing Insight Sept.-Oct. 2015: n. pag. American Marketing Association. American Marketing Association, Sept.-Oct. 2015. Web. 11 Nov. 2015. <https://www.ama.org/publications/MarketingInsights/Pages/channeling-billy-beane.aspx>.
[2] Ibid
[3] Patil, Arun. “North America Text Analytics Market Is Expected to Reach a Value of $1,995.8 Million by 2019 According to New Research Report.” WhaTech. WhaTech, 21 Oct. 2015. Web. 11 Nov. 2015. <https://www.whatech.com/market-research/it/102919-north-america-text-analytics-market-is-expected-to-reach-a-value-of-1-995-8-million-by-2019-according-to-new-research-report>.
[4] Halper, Fern, Marcia Kaufman, and Daniel Kirsh. “Text Analytics: The Hurwitz Victory Report.” (n.d.): 1-22. SAS. Hurwitz & Associates, 2013. Web. 12 Nov. 2015. <https://www.sas.com/news/analysts/Hurwitz_Victory_Index-TextAnalytics_SAS.PDF>.
[5] Ibid
[6] Ibid
[7] Pickett, Stephen. “How Understanding Unstructured Data Is Useful for Customer Insight.” Digital Marketing Magazine. Digital Marketing Magazine, 7 July 2015. Web. 11 Nov. 2015. <https://digitalmarketingmagazine.co.uk/digital-marketing-data/how-understanding-unstructured-data-is-useful-for-customer-insight/2198>.
[8] Patil, Arun. “North America Text Analytics Market Is Expected to Reach a Value of $1,995.8 Million by 2019 According to New Research Report.” WhaTech. WhaTech, 21 Oct. 2015. Web. 11 Nov. 2015. <https://www.whatech.com/market-research/it/102919-north-america-text-analytics-market-is-expected-to-reach-a-value-of-1-995-8-million-by-2019-according-to-new-research-report>.
[9] “Text Analysis; 10 Business Use Cases You May Not Have Thought Of….” Text Analysis Blog. AYLIEN, 19 Aug. 2014. Web. 12 Nov. 2015. <https://blog.aylien.com/post/95184867153/text-analysis-10-business-use-cases-you-may-not>.
[10] Halper, Fern, Marcia Kaufman, and Daniel Kirsh. “Text Analytics: The Hurwitz Victory Report.” (n.d.): 1-22. SAS. Hurwitz & Associates, 2013. Web. 12 Nov. 2015. <https://www.sas.com/news/analysts/Hurwitz_Victory_Index-TextAnalytics_SAS.PDF>.
[11] Turner, Michelle. “Channeling Billy Beane.” Editorial. Marketing Insight Sept.-Oct. 2015: n. pag. American Marketing Association. American Marketing Association, Sept.-Oct. 2015. Web. 11 Nov. 2015. <https://www.ama.org/publications/MarketingInsights/Pages/channeling-billy-beane.aspx>.
[12] Davies, Andrew. “Why Unstructured Data Holds the Key to Understanding the Customer.” Why Unstructured Data Holds the Key to Understanding the Customer. Sift Media, 6 Apr. 2015. Web. 11 Nov. 2015. <https://www.mycustomer.com/feature/data-technology/unstructured-data-key-understanding-customer/169317>.
[13] Turner, Michelle. “Channeling Billy Beane.” Editorial. Marketing Insight Sept.-Oct. 2015: n. pag. American Marketing Association. American Marketing Association, Sept.-Oct. 2015. Web. 11 Nov. 2015. <https://www.ama.org/publications/MarketingInsights/Pages/channeling-billy-beane.aspx>.
[14] Ibid
[15] Halper, Fern, Marcia Kaufman, and Daniel Kirsh. “Text Analytics: The Hurwitz Victory Report.” (n.d.): 1-22. SAS. Hurwitz & Associates, 2013. Web. 12 Nov. 2015. <https://www.sas.com/news/analysts/Hurwitz_Victory_Index-TextAnalytics_SAS.PDF>.
[16] Pickett, Stephen. “How Understanding Unstructured Data Is Useful for Customer Insight.” Digital Marketing Magazine. Digital Marketing Magazine, 7 July 2015. Web. 11 Nov. 2015. <https://digitalmarketingmagazine.co.uk/digital-marketing-data/how-understanding-unstructured-data-is-useful-for-customer-insight/2198>.
[17] Turner, Michelle. “Channeling Billy Beane.” Editorial. Marketing Insight Sept.-Oct. 2015: n. pag. American Marketing Association. American Marketing Association, Sept.-Oct. 2015. Web. 11 Nov. 2015. <https://www.ama.org/publications/MarketingInsights/Pages/channeling-billy-beane.aspx>.
[18] Turner, Vernon, John F. Gantz, David Reinsel, and Stephen Minton. “The Digital Universe of Opportunities: Rich Data and the Increasing Value of the Internet of Things.” EMC Digital Universe with Research & Analysis by IDC. EMC, Apr. 2014. Web. 11 Nov. 2015. <https://idcdocserv.com/1678>.
[19] Ibid
[20] Patil, Arun. “North America Text Analytics Market Is Expected to Reach a Value of $1,995.8 Million by 2019 According to New Research Report.” WhaTech. WhaTech, 21 Oct. 2015. Web. 11 Nov. 2015. <https://www.whatech.com/market-research/it/102919-north-america-text-analytics-market-is-expected-to-reach-a-value-of-1-995-8-million-by-2019-according-to-new-research-report>.
[21] Lamont, Judith. “Text Analytics: Greater Usability, Less Time to Insight.” KMWorld Magazine. Information Today, 29 Oct. 2015. Web. 11 Nov. 2015. <https://www.kmworld.com/Articles/Editorial/Features/Text-analytics-greater-usability-less-time-to-insight-107036.aspx>.
[22] Turner, Michelle. “Channeling Billy Beane.” Editorial. Marketing Insight Sept.-Oct. 2015: n. pag. American Marketing Association. American Marketing Association, Sept.-Oct. 2015. Web. 11 Nov. 2015. <https://www.ama.org/publications/MarketingInsights/Pages/channeling-billy-beane.aspx>.
[23] Pickett, Stephen. “How Understanding Unstructured Data Is Useful for Customer Insight.” Digital Marketing Magazine. Digital Marketing Magazine, 7 July 2015. Web. 11 Nov. 2015. <https://digitalmarketingmagazine.co.uk/digital-marketing-data/how-understanding-unstructured-data-is-useful-for-customer-insight/2198>.
[24] Davies, Andrew. “Why Unstructured Data Holds the Key to Understanding the Customer.” MyCustomer. Sift Media, 6 Apr. 2015. Web. 11 Nov. 2015. <https://www.mycustomer.com/feature/data-technology/unstructured-data-key-understanding-customer/169317>.
[25] Pickett, Stephen. “How Understanding Unstructured Data Is Useful for Customer Insight.” Digital Marketing Magazine. Digital Marketing Magazine, 7 July 2015. Web. 11 Nov. 2015. <https://digitalmarketingmagazine.co.uk/digital-marketing-data/how-understanding-unstructured-data-is-useful-for-customer-insight/2198>.
[26] Keylock, Matt. “The Unstructured Data Challenge.” Dunnhumby. Dunnhumby, 12 Dec. 2012. Web. 12 Nov. 2015. <https://www.dunnhumby.com/insight/the-unstructured-data-challenge>.
[27] Patterson, Laura. “Why Your Data Scientists Need to Be Storytellers, and How to Get Them There.” MarketingProfs. MarketingProfs LLC, 12 Nov. 2014. Web. 11 Nov. 2015. <https://www.marketingprofs.com/articles/2014/26436/why-your-data-scientists-need-to-be-storytellers-and-how-to-get-them-there>.
[28] Turner, Vernon, John F. Gantz, David Reinsel, and Stephen Minton. “The Digital Universe of Opportunities: Rich Data and the Increasing Value of the Internet of Things.” EMC Digital Universe with Research & Analysis by IDC. EMC, Apr. 2014. Web. 11 Nov. 2015. <https://idcdocserv.com/1678>.
[29] Halper, Fern, Marcia Kaufman, and Daniel Kirsh. “Text Analytics: The Hurwitz Victory Report.” (n.d.): 1-22. SAS. Hurwitz & Associates, 2013. Web. 12 Nov. 2015. <https://www.sas.com/news/analysts/Hurwitz_Victory_Index-TextAnalytics_SAS.PDF>.
[30] Lamont, Judith. “Text Analytics: Greater Usability, Less Time to Insight.” KMWorld Magazine. Information Today, 29 Oct. 2015. Web. 11 Nov. 2015. <https://www.kmworld.com/Articles/Editorial/Features/Text-analytics-greater-usability-less-time-to-insight-107036.aspx>.
[31] Ibid
[32] Davies, Andrew. “Why Unstructured Data Holds the Key to Understanding the Customer.” MyCustomer. Sift Media, 6 Apr. 2015. Web. 11 Nov. 2015. <https://www.mycustomer.com/feature/data-technology/unstructured-data-key-understanding-customer/169317>.
[33] Ibid
[34] Burns, Ed. “How The New York Times Uses Predictive Analytics Algorithms.” SearchBusinessAnalytics. TechTarget, Oct. 2015. Web. 13 Nov. 2015. <https://searchbusinessanalytics.techtarget.com/feature/How-The-New-York-Times-uses-predictive-analytics-algorithms>.
[35] Lamont, Judith. “Text Analytics: Greater Usability, Less Time to Insight.” KMWorld Magazine. Information Today, 29 Oct. 2015. Web. 11 Nov. 2015. <https://www.kmworld.com/Articles/Editorial/Features/Text-analytics-greater-usability-less-time-to-insight-107036.aspx>.
[36] Turner, Michelle. “Channeling Billy Beane.” Editorial. Marketing Insight Sept.-Oct. 2015: n. pag. American Marketing Association. American Marketing Association, Sept.-Oct. 2015. Web. 11 Nov. 2015. <https://www.ama.org/publications/MarketingInsights/Pages/channeling-billy-beane.aspx>.
[37] Ibid.
[38] Davies, Andrew. “Why Unstructured Data Holds the Key to Understanding the Customer.” MyCustomer. Sift Media, 6 Apr. 2015. Web. 11 Nov. 2015. <https://www.mycustomer.com/feature/data-technology/unstructured-data-key-understanding-customer/169317>.
[39] Keylock, Matt. “The Unstructured Data Challenge.” Dunnhumby. Dunnhumby, 12 Dec. 2012. Web. 12 Nov. 2015. <https://www.dunnhumby.com/insight/the-unstructured-data-challenge>.
[40] Turner, Michelle. “Channeling Billy Beane.” Editorial. Marketing Insight Sept.-Oct. 2015: n. pag. American Marketing Association. American Marketing Association, Sept.-Oct. 2015. Web. 11 Nov. 2015. <https://www.ama.org/publications/MarketingInsights/Pages/channeling-billy-beane.aspx>.
[41] Ibid
[42] Pickett, Stephen. “How Understanding Unstructured Data Is Useful for Customer Insight.” Digital Marketing Magazine. Digital Marketing Magazine, 7 July 2015. Web. 11 Nov. 2015. <https://digitalmarketingmagazine.co.uk/digital-marketing-data/how-understanding-unstructured-data-is-useful-for-customer-insight/2198>.
[43] Turner, Michelle. “Channeling Billy Beane.” Editorial. Marketing Insight Sept.-Oct. 2015: n. pag. American Marketing Association. American Marketing Association, Sept.-Oct. 2015. Web. 11 Nov. 2015. <https://www.ama.org/publications/MarketingInsights/Pages/channeling-billy-beane.aspx>.
[44] Lamont, Judith. “Text Analytics: Greater Usability, Less Time to Insight.” KMWorld Magazine. Information Today, 29 Oct. 2015. Web. 11 Nov. 2015. <https://www.kmworld.com/Articles/Editorial/Features/Text-analytics-greater-usability-less-time-to-insight-107036.aspx>.
[45] “Why Marketers Still Haven’t Mastered Personalization.” EMarketer. EMarketer Inc., 23 Sept. 2015. Web. 12 Nov. 2015. <https://www.emarketer.com/Article/Why-Marketers-Still-Havent-Mastered-Personalization/1011220>.
[46] Keylock, Matt. “The Unstructured Data Challenge.” Dunnhumby. Dunnhumby, 12 Dec. 2012. Web. 12 Nov. 2015. <https://www.dunnhumby.com/insight/the-unstructured-data-challenge>.
[47] Ibid
[48] “Why Marketers Still Haven’t Mastered Personalization.” EMarketer. EMarketer Inc., 23 Sept. 2015. Web. 12 Nov. 2015. <https://www.emarketer.com/Article/Why-Marketers-Still-Havent-Mastered-Personalization/1011220>.