How Analytics-Savvy do Marketers Need to Be?

How analytics savvy do marketers need to be?
Today, we continue our discussion on Big Data and Analytics. Check out our previous posts on the topic, Big Data: What is it & How Can it be Used? and 8 Things Marketers are Doing with Big Data and Analytics.

Marketers are not expected to build complex models to analyze results; however, we are expected to be data and analytics savvy.

This includes:

  • understanding what analytics can and cannot achieve
  • knowing the necessary questions to ask when designing and evaluating a project
  • having a basic comprehension of the math and statistics that underlie models and are used to evaluate their validity
  • being confident about making decisions based upon the output

Our longer-term goal is to move essential, everyday analyses into the hands of marketers and line-of-business managers so that we can act quickly on marketplace developments. To do so requires that complex models be transformed into highly interactive tools, with graphics-based user interfaces that make it easy for laypeople to visualize, understand, and further explore data. As this becomes a reality, and data visualization tools are making it possible even today, marketers will need to get increasingly comfortable with conducting basis analytics.

Toward this end, a great resource for marketers is Keeping Up with the Quants, Your Guide to Understanding + Using Analytics, a book coauthored by Tom Davenport and Jinho Kim.18

Tackling an Analytics Project

A considerable amount of upfront work is required to ensure the success of any data-analytics project. Experience shows that organizations that spend enough time upfront developing a clear sense of what they want to achieve have considerably more success in the end.

One of our first considerations is identifying and garnering buy-in from the key decision makers related to the challenge being addressed. By involving key stakeholders in the process, we can improve our chances of understanding the underlying challenge and having access to the necessary data and resources to effectively address it. Understanding and managing stakeholders’ expectations for the project and agreeing early on ways to provide regular feedback help ensure that our findings are implemented.

Essential questions for the stakeholder group to ask at the onset of the project include, “What is the problem we are trying to address?” and “How will this undertaking create business value?” In the absence of clarity about these questions, even the most sophisticated analytics will amount to a muddle. To get at these answers, companies often engage in destination thinking. A simple process, destination thinking encourages us to start with the end in mind by writing down the business impact we are trying to achieve in short, clear sentences. The goal should be defined as specifically as possible because analytics require clear parameters.

Davenport encourages us to investigate any prior research that has been conducted addressing this challenge.

  • How have these projects been designed in the past?
  • How have they defined the problem?
  • What types of data have been used to answer the question?
  • How might we approach the question from a different angle that might generate new insight?
  • What other parameters are important for creating successful data and analytics solutions?

Jesse Harriott, chief analytics officer for the online marketing company Constant Contact and Jean Paul Isson, global vice president of business intelligence and predictive analytics at the employment company, Monster Worldwide, Inc., have overseen many analytics challenges during the course of their careers. In their book, Win With Advanced Business Analytics, they provide simple and useful guidelines for translating business challenges into relevant and manageable analytics projects.

To offer parameters for creating successful analytics solutions, Harriott and Isson adapted the tried and true SMART framework,19 originally developed by management consultant George Doran. The adapted SMART framework for successful analytics solutions is:

  • Specific: Having a clear mandate and goal, defined by clear, expected benefits.
  • Measurable: We can easily track performance.
  • Aggressive and Actionable: A solution that could be quickly put into the organization and that leads to clear action.
  • Realistic: A solution that avoids over promising or overestimating the value added of the outcome.
  • Time bounded: The solution should have a clear timeline to deliver the business benefit to the organization.

Build the Model

Models are representations of the underlying challenge we are aiming to solve. Analytical models are essentially math equations that describe relationships among the variables that we are looking to evaluate. An art/science modeling also involves understanding which datasets and variables have a predictive influence on the challenge at hand and which analytical techniques are the most applicable. Modelers benefit from a solid understanding of data science and business as this combination of expertise helps modelers build successful models and solve the black box problem—business people not trusting or understanding the output of the model—because they can explain why the underlying model accurately addressing the business challenge at hand. Even when working with best modelers, marketers still need to satisfy our understanding of the model by inquiring about the assumptions, the causality between variables, and the data—its source, how it may have been enriched, and in the case of data fusion, the strength of the common characteristics between populations—in order to trust its output.

What types of analytical techniques are being used to address marketing-related business challenges? In their book Win With Advanced Business Analytics, Harriott and Isson created an Analytics Recipe Matrix, an adaptation of which appears below, that provides a sampling of the types of challenges that can be tackled with analytics.

Analytics Recipe Matrix

Find the Data

The next challenge to tackle is identifying what data sources will provide the necessary inputs and how we can access them. It may be hard to believe that given the amount of data that is out there, there may still be holes in our data; however, this is not uncommon. Sometimes data may exist, but we cannot get to it. An example might be the data generated by a new social new platform for which analytics have not yet been developed. Other times data may exist, but it is in a format that we are not yet able to manipulate.

Much of the data that our organizations have is proprietary—it comes from our own data streams. That does not necessarily mean that we have ready access to it, however. Data is often captured by business function; web analytics are often managed separately from customer-service metrics, marketing-campaign analysis, and sales-force productivity figures. Indeed, according to a survey conducted by the CMO Council, 52 percent of marketers list functional silos as top hindrances to customer-centric endeavors.

52 percent of marketers list functional silos as top hindrances to customer-centric endeavors Click To Tweet

Further, systems often define terms and measure differently. Analytic modelers are necessary to collect, clean, standardize, and often enrich data so that it can be effectively analyzed. This is often the most time-consuming component of the entire analytic project because of the number of details. However, it is key to generating worthwhile insights: garbage in = garbage out.

Deploy the Model

If we have done the upfront work, deploying the model is relatively straightforward. To maximize the impact of any analytic effort, Davenport suggests that we operationalize the model. For example, a churn analysis can be turned into a daily dashboard for sales reps that flags customers who are displaying behaviors commonly associated with defection. Whenever we can, we should encourage our modelers to transform their work into user-friendly decision-making tools that can be readily employed by those who are charged with transforming analytic insights into action.

Put Insights to Work and Measure the Impact

The final steps in any analytics undertaking are to put the insights to work and measure the impact of the decision on business results. What have we learned? What will we do next as a result? Sharing our results with the broader enterprise facilitates learning and buy-in for further analytic endeavors. This can be accomplished through a variety of ways, including a formal communication plan, a knowledge portal, and even Lunch-and-Learn discussions. Keeping the C-suite abreast of new learning is key to our ability to ultimately transform data into a strategic asset.

Develop a Marketing Data Analytics Capability

Developing a sophisticated data analytics capability in-house is fundamental to our ability to offer truly customer-centric experiences. While we can outsource this capability, it is in our best interest to grow this competency internally, to have it be part of our ongoing workflow as a way we routinely evaluate our business.

Companies’ abilities in this realm currently run the gamut. Davenport devised a five-tier analytics pyramid to categorize organizations based upon their use of big data analytics to impact business results. As seen below, the stages range from analytically impaired to analytical competitors.

How Analytical Is Your Company?

In Davenport’s experience, the majority of companies are currently in stage two, where impact remains localized, not yet shared across an ecosystem. Some highly analytical companies remain at the second tier of the pyramid because their analysis remains based on small datasets. A few companies, mostly smaller companies and a handful of large enterprises like P&G, GE, and Novartis, are at stage three, leading the way, defining what is possible. These companies blend data and analytics seamlessly and apply the insights into running their day-to-day businesses. Viewing data as a strategic asset, analytical tools are available at the point of decision, redefining agility, and applying insights at scale.21

Becoming a top-tier analytical competitor is complicated as it involves people with a range of understanding and commitment to data and analytics, complex and disparate systems, and myriad processes and policies. Several hurdles must be crossed in order for data and analytics to grab a hold and prove results. As complicated as finding insight in big data may be, it is not an impossible task. Bill Franks, chief analytics officer for Teradata and author of Taming the Big Data Tidal Wave, encourages us to keep perspective. Big data, according to Franks, is “simply a continuation of the struggle we’ve always had to incorporate ever-growing and ever more diverse data sources into analytics to enable better business decisions.” Besides we do not have to get there overnight. Companies are making progress by taking an incremental approach to adoption.

Think big. Start small. Scale fast.

Big Data Means Big Responsibility

Given the sensitive nature of much of the data that we collect on our customers, we have to take great care to not turn this opportunity to deliver enhanced customer-centricity into a disaster.

Our customers trade off privacy for customized interactions. Although they are not always aware of the connection, our customers cannot have websites that greet them by name and offer personalized recommendations without small text files, or cookies, being left on their computers that facilitate this memory, e-mail addresses being captured, or having the content they are reading on their screens scanned.

How do our customers feel about their privacy? According to the research conducted by the Economist Intelligence Unit, 49 percent of consumers say they are “very concerned” about the threat of privacy online.23 However, only 23 percent of marketers believe that their customers are concerned about privacy. The increase in apps that thwart location identifiers and interject random data into people’s data-streams, along with the popularity of ephemeral social networks like SnapChat, where content disappears unless captured in a screenshot, suggest that people are increasingly concerned and taking matters into their own hands. If we are smart, we will proactively give our customers reason to trust us in this realm by making data governance an essential component of our customer experience. A new field, data governance focuses on how we can best manage data across enterprises and ecosystems. It’s scope includes the technology that gathers, integrates, cleans, sorts, and analyzes data, as well as the people, policies, and processes that surround it. Data governance also includes preventing and managing security breaches. It helps our customers know that their data is safe with us and ensures that the data from which we derive our insights is trustworthy. Neither is trivial.

As we create data policies Bill Franks cautions us to consider the trifecta of what is legal, ethical, and acceptable to the public. In blog post published by the International Institute for Analytics, he warns that in this rapidly changing field, laws may lag our data collection and analytic capabilities and customer sensitivities may shift.24 Franks recommends that we take a cautious approach, pursuing strategies that meet all three considerations simultaneously. Citing a large U.S. superstore chain’s experience in predicting the early-stage pregnancy of its customers based on their purchase history, a summary of which follows, Franks cautions that what is legal and ethical may still not be acceptable to our customers.

Marketing Gets a Little Too Personal

One large data-savvy superstore chain has experienced success targeting its customers with offers based upon a number of behaviors including past purchases. A strong correlation was discovered between pregnancy and purchases of items such as unscented lotion and soap, vitamin supplements, and extra-big bags of cotton balls. So reliable were these predictions, that the company was able to estimate the mother’s due date within a fairly small window. Operating on this finding, the retailer made it a practice to send coupons for baby items to customers who were flagged as potentiality pregnant.

One of the variables the company did not consider in their analysis was the age of the customers it was targeting. In one instance, the recipient of the coupon mailing happened to be a teenager who was living at her parents’ home. When she received the coupons, the young woman’s father became irate, stormed into their local superstore, and asked the manager, “Are you trying to encourage her to get pregnant?” The manager apologized profusely. Several days later the manager called the father again to apologize. This time the father had a different tone. “I had a talk with my daughter,” he said. “It turns out there’s been some activities in my house I haven’t been completely aware of. She’s due in August. I owe you an apology.”25 Despite the accuracy of its targeting and its good intentions, the superstore chain had crossed a boundary.

Taking Their Data Back

Recognizing the value of their data, coupled with concerns about their privacy, our prospects and customers may become more selective about what they will share with us in the future. To continue to have access to their digital footprint, companies will have to provide real value in return. Being radically transparent with our customers about the link between access to their data and our ability to provide them with a quick, easy, and personalized shopping experience is an important step.

As marketers we also need to be transparent about how we collect data and how we will and will not use it with our customers. Facebook is very clear about its view on privacy; it refers to it as a Data Usage Policy. Acxiom, which supplies marketing data to businesses, provides anyone with the opportunity to see what data is being collected about them and how it is being used by advertisers. Allowing customers who prefer anonymity to opt out of our marketing efforts will help them feel respected and more in control. Offering tools that allow customers to proactively share additional data to further enhance their experience–such as Amazon’s Improve Your Ratings feature–helps customers on the other end of the spectrum feel respected and in control.26

Doc Searls, author of The Intention Economy: When Customers Take Charge, envisions a future in which power shifts even more dramatically toward our customers.27 He anticipates a time in the not-too-distant future when people actively control the flow and use of their personal data, dictate their own terms of service, and tell the marketplace what they want, how they want it, where and when they want it, and how much it should cost.

Under this scenario, an individual could send a Request for Proposal to a company or group of companies or the entire marketplace outlining the parameters upon which they would rent a car, purchase a house, or buy insurance. This announcement could take place anonymously or publicly and would specify the time frame for his or her willingness to receive information from the vendors—say July and August in anticipation of a late August purchase. This brings a whole new meaning to our notion of permission-based marketing!

Searls calls this the Intention Economy and believes that it will more efficiently link supply and demand than the system we currently have in place. Rather than customer relationship management systems, vendor relationship management (VRM) tools, technologies, and services will form that will allow people to manage their unique identities—imagine only having to input a change of address once and have it reach multiple vendors—as well as their demand. Vendors that make use of these tools will be able to build better relationships with their customers and close sales. Until such time as our customers are able to speak so clearly for themselves, we have data analytics to help us understand what makes them tick.

QUESTIONS TO CONSIDER

What stage of the five-tier analytics pyramid describes your company?
What data is your company currently collecting? Who owns it, manages it, and is responsible for its security?
What questions would you like to be able to answer if you were not limited by your current data capabilities?
What is your next step?

This content is an example of what ContentOro does for its customers…providing high-quality, relevant content from experts and their published books.

About the Authors

Larry Weber and Lisa Leslie Henderson are the co-writers of this Digital Marketing guide. Larry is the CEO of Racepoint Global, an advanced marketing services firm. A globally known expert in public relations and marketing services, Larry has successfully built companies and brands and is passionate about the future of marketing. Lisa is an observer, synthesizer, and writer who draws extensively from her background in marketing and consulting. Lisa and Larry have collaborated on two guides to date, The Digital Marketer, and Everywhere: Comprehensive Strategy for the Social Media Era. To stay current on their thinking, frequent www.racepoint.com/thedigitalmarketer and follow them at @TheLarryWeber and @ljlhendo.

Buy on Amazon: The Digital Marketer: Ten New Skills You Must Learn to Stay Relevant and Customer-Centric

18. Thomas Davenport and Jinho Kim, Keeping Up with the Quants, Your
Guide to Understanding + Using Analytics . (Boston: Harvard Business
Review Press).
19. Jean Paul Isson and Jesse Harriott, Win With Advanced Business Analytics
Analytics (New Jersey: John Wiley & Sons, Inc.), 47–48.
21. Thomas Davenport, “Analytics 3.0: Measuring Business Impact from Analytics
and Big Data,” Webinar, Harvard Business School, www.slideshare.net/boscolg/analytics-30measurable-business-impact-from-analytics-big-data
23. Lyris, “Mind the Digital Marketing Gap: New Findings by Lyris & the
Economist Intelligence Unit,” (infographic), Lyris Connections Blog, https://visual.ly/community/infographic/business/mind-digital-marketing-gap-%E2%80%93-new-findings-economist-intelligence-unit
24. Bill Franks, “Helpful or Creepy? Avoid Crossing the Line with Big Data,”
International Institute for Analytics , http://iianalytics.com/2013/05/helpful-or-creepy-avoid-crossing-the-line-with-big-data/
25. Kashmir Hill, “How Target Figured Out a Teen Girl Was Pregnant Before
Her Father Did,” Forbes , www.forbes.com/sites/kashmirhill/2012/02/16/how-target-figured-out-a-teen-girl-was-pregnant-before-her-father-did/
26. For more information see: www.amazon.com/gp/help/customer/display.html/?nodeId= 13316081
27. “VRM Vision,” Project VRM , http://cyber.law.harvard.edu/projectvrm/VRM_vision

ContentOro
ContentOro revolutionizes the way digital marketers and brands acquire content for marketing. We partner with the world’s best publishers to bring the contents of their books to life on the web with our innovative technology. Creating compelling experiences that tie our content to our clients’ product is our mission.

Leave a Reply Text

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