CHAPTER 1

Why Has Analytics Missed the Mark?

Chapter Overview

Why has analytics missed the mark? This chapter first explores the main reasons responsible for the lackluster performance of marketing analytics and then introduces three prevailing analytics maturity models that not only will help organizations to recognize their analytic maturity state but also provide macrolevel solutions to achieving analytic successes.

This chapter is organized as follows:

• Why Has Analytics Not Yet Lived Up to the Promise?
• Davenport’s Five-Stage Analytics Maturity Model
• Davenport’s DELTA Model
• The Gartner Continuum Model
• Wayne Eckerson’s Analytical Maturity Model
• Challenges and Opportunities Facing Small- to Medium-Sized Companies
• The ALADA Model—Five Pillars of Analytics Success for Small- to Medium-Sized Firms
• Conclusion

Why Has Analytics Not Yet Lived Up to the Promise?

Analytics is the scientific process of transforming data into insight for making better decisions. Since Professor Tom Davenport published his breakthrough book—Competing on Analytics: The New Science of Winning, more leaders see analytics as a new wave of competitive advantage. The application of analytics is becoming commonly accepted in marketing.

Investments in analytics have been increasing steadily. The 2018 chief marketing officer (CMO) Survey conducted by Duke University’s Fuqua School of Business reports that the percentage of marketing budgets companies plan to allocate to analytics over the next three years will increase from 5.8 percent to 17.3 percent, a whopping 198 percent increase.1

However, in the same CMO Survey, top marketers also report that the effect of analytics on companywide performance remains modest, with an average performance score of 4.1 on a seven-point scale, where 1 = not at all effective and 7 = highly effective. It is bothersome that the analytics performance impact had shown little increase over the last five years when it was rated 3.8 on the same scale.2

A 2015 Forbes Insights Report indicates that only 22 percent of marketers have data-driven initiatives achieving significant results. According to ITSMA and Vision Edge Marketing, 74 percent of marketers can’t measure or report how their efforts impact their business.3 These numbers are startling, considering the importance and prevalence of analytics in marketing success and proving return on investment (ROI).

So, why has analytics missed the mark? Below are eight of the most common reasons:

Data Problems

Data make or break a business because that data fuel marketing analytics. Commonly seen issues with data are as follows:

1. Poor data quality. Poor quality data include inconsistent data, missing data, wrong data, duplicate data, and outdated data. Several reasons for poor data quality include:
   • Lack of budget for timely data merge and hygiene. For instance, one customer may have multiple records in the database under different names;
   • Outdated store POS system that is unable to capture key customer information;
   • Human error. For instance, sales rep typed the name wrong into the database;
   • Data value is not consistent across all databases because IT only updates selected databases;
   • External data feeds were not imported into the databases promptly;
   • Data dictionary was created by IT, not by the businesspeople.
2. Scattered and disconnected data. Data are typically owned and maintained in separate systems by separate departments across organizational silos. There is(are) no common variable(s) that can stitch them together. For instance, the CRM database, social, e-commerce, and call center data are stored in different databases, and they disconnect from each other.
3. Inaccessibility to data. Data are not available to all stakeholders. Because data are isolated in different systems and places, marketers cannot access some of the critical pieces of information about customers. A typical example is that e-commerce, call center, and marketing team are three separate business units, and the e-mail marketing team typically does not have access to the CRM database, and vice versa.
4. Insufficient data breadth. Many people say we are living in an era of big data overflow. While companies do seem to have far more data than they can process, the reality is that due to budget constraints and technical difficulties, they do not have enough useful data that can be leveraged for analytics and action. Useful data include both structured data and unstructured data. The structured data are data stored in a relational database such as customer demographics, behaviors, campaign responsiveness, product usage, cross-channel interaction, and so on; the unstructured data are data that aren’t stored in a fixed record length format. Examples include documents, social media feeds, digital pictures and videos, call center interactions, on-site interactions, survey opinions, and so on. Lack of data breadth limits an organization’s ability to gain deeper insights into its customers.
5. Not using external data. There are two reasons why some companies are not taking advantage of external data. First is lack of budget. Companies, especially small- to medium-sized companies, do not have a budget for purchasing external data such as customer demographic, geographic, and attitudinal information. Second, although there are so many data (i.e., economic, job, population, weather, housing, etc.) available free to the public that can be used for research and modeling, some analysts are either not aware of them or do not know where to find them from public domains.
6. Poor data management and governance. Many companies do not have an effective data governance strategy. There are no good QA and QC procedures in place to ensure the integrity of the data. Data dictionary was not created or updated promptly. Data processing procedures were not properly written and archived; knowledge got lost in transitions after key personnel left or because of the change of service providers.

Technology and IT Support Issues

Some commonly seen problems include but are not limited to:

• Outdated and rigid legacy data system. For instance, the outdated store POS system is unable to store some key customer data. Replacing such a system requires a lot of money. In some cases, the database is so old that the database administrator dares not make any changes to tables for fear that any significant changes will trigger a collapse of the entire database system.
• Lack of IT support. IT team does not allocate enough people to support the marketing and analytics team.
• Lack of effective communications. Marketing treats IT as a back-office function. There is no consistent and meaningful communication between IT and marketing, marketing and analytics, and analytics and IT. For instance, marketing and analytics decided to purchase a new analytics software without consulting IT. Later, they found out that additional servers are required to host it, but neither the marketing nor the IT side had the extra budget for purchasing the servers. That analytical software ended up sitting idle for several months until marketing secured the additional funds in the next fiscal year.

Poor Investment Decisions

Poor investment decisions waste your limited marketing dollars, which is one of, if not, the biggest reasons why your analytics ROI was unsatisfying.

• Assuming the wrong approach to tool and software selections. Companies with low analytics IQ tend to choose tools and software not based on their capabilities but rather based on how the vendors claim their tools can solve companies’ primary problems. Often, these companies do not have proper protocols, standards, and procedures in place to compare and evaluate tools from different vendors. Therefore, they cannot objectively compare the pros and cons of each vendor and make the right purchase decisions.
• Buying vanity software and tools that add little value to the business. A good example is the multichannel campaign management system, which could easily cost retailers half a million dollars every year. For instance, many retailers believe that a campaign management system is a must-have to pull campaigns, but indeed, it is not, especially for companies that do not have many customer records. You may use free tools such as R or the Basic SAS to create an in-house campaign management system that will cost you nothing or only cost several thousand dollars a year. The in-house campaign management system not only can save you up to quarter million dollars a year, it will also significantly improve the work efficiency and shorten the turnaround time of campaign creation by at least a day or two.
• Overspending on tools and data not on people. Some companies spend more than 80 percent of their analytics budget on tools and data, and less than 20 percent on people and training. They tend to buy tools and technologies that have way more functions than what they need. For instance, data vendors like to pitch two sexy things to marketers: First, data must be comprehensive, which means that to develop a true 360-degree customer view, you must capture every touch point and every nuance of a customer crossing the entire customer journey. Second, you must capture data in real time so you can respond to their behavior promptly. Both sound very attractive and seem to make perfect sense. But the problem is that capturing every touch point is highly expensive and arguably unattainable. Data are only as good as you use it. If you don’t use it, you waste your time and money big time.
• Letting IT, not the business, lead the software search. IT project managers usually do not quite understand what the business side wants. They tend to choose products based on technical requirements rather than business requirements.
• Not getting IT involved when selecting tools. When selecting analytical tools, ignoring IT partners’ opinions or even totally not getting IT involved will cause problems, too. For instance, the analytics team of one of my clients decided to purchase an analytical tool without consulting the IT team; only later, they found out that the IT side lacks the knowledge to maintain and support that device.
• Falling into the trap of “User-Friendly.” Software vendors like to pitch user-friendly features to executives, and that trick always works like a charm. Don’t get me wrong; we all want user-friendly tools. The problem is that more often than not, you get these so-called user-friendly features at the expense of sacrificing much-needed functionalities and flexibilities. Even worse, many so-called user-friendly tools are not so user-friendly at all. Also, a genuinely user-friendly tool does not always guarantee that users will use it. For instance, many companies purchased the expensive Tableau viewer licenses for their executives, hoping that they would pull reports every day by themselves. However, what we’ve found out was that only a small number of executives would do that. Most executives still prefer the marketing or analytics team pulling the reports and presenting them the insights instead of doing the job by themselves.

Lack of Analytics Marketers

To compete on analytics, companies desperately need to bring the left-side and the right-side brains together. The convergence of marketing and analytics, which was once nice to have, is now becoming a new trend and business-critical. Analytics marketers are individuals who know analytics and can also speak the language of business. They translate business requirements into terms that analytics and technologists can understand. Conversely, they can also use plain English to show business the value of data, justify investments in analytics, and translate insights gleaned from data into easy-to-understand stories for better decision making. Analytical marketers are critical to improving the competitive advantages of the company. The more analytics marketers a company has, the more likely the company is to adopt an analytically oriented culture and use analytics to make better decisions than the competition. Many companies do not have qualified analytics marketers; they are not yet ready for the era of insight-driven marketing.

Lack of Executive Support

Like any other projects, support from executives is critical to the success of analytics. Organizations need analytical leaders to set and clarify strategic objectives and ensure appropriate project funding. Analytical leaders help secure resources, provide project governance, create high-level organizational buy-in from all stakeholders, manage risks, and make critical decisions. Leading analytically oriented companies often have a couple of senior executives sponsor analytics initiatives; they raise the awareness and analytics IQ within the organization and create and maintain a culture of excellence.

Scarcity of Analytics Professionals and Skills

The shortage of analytics skills means several things:

1. Lack of experienced analytics professionals. An analytics team consists of several types of analytics professionals, such as data scientists, data developers, statisticians, and business analysts. As more organizations embrace artificial intelligence and machine learning technologies to achieve competitive advantages, good analytics people, especially those who have deep knowledge in data and statistics and also have excellent SAS/R/Python programming skills, are in high demand. Finding the right analytics talent is challenging for all companies.
2. Lack of business domain knowledge and experience. Most analytics professionals come from math and statistics background and have little or no business domain knowledge. Therefore, some of the “smart insights” they discover have no real business benefits. Some analysts can only find patterns but are unable to form assumptions and hypotheses to determine the root causes further.
3. Lack of innovation in analytical techniques and methodologies. Many organizations keep using approaches and methods that once worked very well and don’t try to improve them or don’t know how to improve them. For instance, in the past 10 years, response rates and ROIs of many traditional marketing channels such as direct mail, print catalog, and e-mail have plateaued or even declined year-over-year. One of the reasons, of course, is that marketers’ attention and resources have shifted to mobile and online marketing where marketers see more growth opportunities. However, lack of innovation in analytical techniques and the slowness of adopting new technologies such as machine learning are two major reasons that have created the low effectiveness of marketing initiatives.
4. Poor communication skills. Some analysts cannot use easy-to-understand English to effectively communicate the findings, insights, and actionable recommendations to marketers.

Company Silos

The company silos present when certain departments or sectors do not wish to share knowledge and information with others in the same company. This lack of information flow results in departmental isolation and territorialism. Company silos can lead to negative customer experience, inefficiency, and duplicate work, and may also contribute to the demise of productive data-driven company culture.

• Negative Customer Experience. Negative customer experience is often caused by either operational silo, channel silo, or both. The operational silo happens when the various business units present in an organization aren’t aware of one another’s operations and decisions and act autonomously without getting insights from related business units. For instance, recently, I called the customer service of a carpet cleaning company to set up an appointment because I received a great e-mail offer from them. The call center rep searched the database and couldn’t find the offer code in the database. Her supervisor has the authority to see more data in the system. She was able to locate the coupon code, but she told me that the e-mail was sent out by another franchise whose service territory is out of the scope of my home; therefore, they couldn’t use the coupon code. I was disappointed. Customers do not care about how your franchise system works; they view all franchisees as one in the same brand. At the very least, a disclaimer should be included in the e-mail to remind recipients that the offers are subjective to specific regions.
• Duplicate Work. In some large organizations, it is not uncommon for each department to have similar positions performing similar functions. Their jobs are highly overlapped. Departmental silos also cause companies to spend money, which is avoidable. For example, I noticed that the credit card department of one of my clients spent a significant amount of money and hired a consulting firm to build a reporting package for them. What I found out was that the company’s analytics team could have developed that reporting package much quicker if the credit card department had reached out to them for help.
• Inconsistent Branding. Today, retailers recognize the importance of omnichannel marketing. An omnichannel strategy focuses on delivering consistent brand presence across channels during a consumer’s buying process and making that buying process a seamless and consistent experience. Channel silos hinder the continuous flow of contextual and historical information between channels, thus resulting in inconsistent brand images and different customer experience across different channels.

Not Using the Right Metrics and Key Performance Indicators

Metrics are quantifiable measures that are used to monitor and evaluate financial performance, reveal the truth about performance, and provide an actionable way to achieve overall business strategies and goals. Key performance indicators (KPIs) are a subset of metrics that provide a simple, insightful snapshot of a company’s overall performance, as well as reliable, real-time information for effective decision making. Continuously tracking the trends of KPIs for an extended period will help highlight any issues that might otherwise go unnoticed and discover hidden opportunities for further growing your business. However, in reality, quite a few companies make decisions without using the right metrics, thus resulting in low ROI of marketing initiatives. For instance, recently, quite a few multichannel retailers repositioned print catalogs as a branding tool to raise brand awareness and drive traffic to other channels. Did all these retailers make the decision based on KPIs such as customer lifetime value and incremental margin? Probably not. A couple of them I knew made that move simply because other retailers did so.

So, how to fix these problems and improve the ROI of your investments in marketing analytics? The first step, of course, is to identify the gap between you and those analytically competitive firms. Below I’d like to introduce you to four analytics maturity models that will help you not only recognize where you are at on the analytics maturity curve but also get you on the right path toward analytical success.

Davenport’s Five-Stage Analytics Maturity Model

In 2007, Thomas H. Davenport and Jeanne G. Harris published their famous Five Stages of Analytics Maturity Model that classifies organizations into five maturity stages based on five success factors: data, enterprise, leadership, targets, and analysts.

Davenport describes five stages of analytics maturity as follows:

1. Analytically impaired: The organization is flying blind and reactive. It is plagued by missing, inconsistent, or poor-quality data, multiple definitions of data, and poorly integrated systems. Leadership is not aware of the importance of analytics or simply has no interest in analytics. Analysts have low analytical skills, and they are usually attached to specific functions.
2. Localized analytics: Data are usable but in functional or process silos. The organization collects transaction data efficiently but often lacks the right data for better decision making. Analytic efforts are isolated, opportunistic, and function specific, lacking communications across different departments and teams. Leadership is only at the functional or process level. Targets are disconnected and may not be strategically important.
3. Analytical aspirations: The organization has a proliferation of business intelligence (BI) tools and data marts, but most data remain unintegrated, nonstandardized, and inaccessible. Organizations begin to create a centralized data repository. Early stages of enterprisewide approaches start appearing. Executives recognize the importance of analytics. Many key areas hire analysts, and analytical efforts coalesce behind a small set of targets.
4. Analytical companies: Leaders embrace analytics and support analytics initiatives. Data quality is high. The organization develops an enterprisewide analytical plan, instills proper IT processes and governance principles, and embeds some automated analytics. They hire highly capable analytics professionals. The organization begins to develop an enterprisewide analytics capability, which is viewed as a corporate priority. Analytical activities center on a few key domains.
5. Analytical competitors: Leaders have a strong passion for analytical competition. They hire the best analytics professionals. The organization relentlessly searches for new data and metrics and has a full-fledged analytic architecture that is enterprisewide, fully automated, and integrated into processes. The organization is routinely reaping big benefits from its enterprisewide analytics capability and focuses on making that business advantage renewable.4

Currently, most organizations are sitting at stage two or three in Davenport’s model. According to a recent study by Gartner, almost 90 percent of organizations have low BI and analytics maturity.5

Davenport’s DELTA Model

In their 2010 book, Analytics at Work: Smarter Decisions, Better Results, Tom Davenport, Jeanne Harris, and Bob Morison introduced the DELTA model. If the Five Stages of Analytics Maturity Model is the industry standard framework for assessing analytics maturity, the DELTA model is the compass for achieving analytics success. As Davenport explains, the DELTA model has five attributes (later, two new elements Technology and Analytical Techniques were added, thus becoming the DELTA Plus model):

• D (Data)—must have large amounts of integrated, high-quality, and easily accessible data about their businesses and markets. Technology-wise, you’ll need heavy-duty hardware and software to do serious analytical work.
• E (Enterprise)—must have an enterprisewide approach. Instead of managing your analytics resources in disconnected silos, highly analytical firms manage these data, technology, and analytics people in a coordinated fashion throughout the enterprise.
• L (Leadership)—must have leadership supporting analytics initiative. One of the key factors driving success in analytics is the strong, committed leaders who understand the importance of analytics and constantly advocate for their development and use in decisions and actions.
• T (Target)—must identify key target projects. The analytical competition requires a clear business strategy that is optimized with data and analysis. Your executives should begin to consider what key processes and strategic initiatives would be advanced if the right analytics were available.
• A (Analyst)—must hire and train high-quality quantitative analysts and data scientists.

The Gartner Continuum Model

In addition to Davenport’s analytics maturity model, two other analytic maturity models are worth your attention—the Gartner Analytic Continuum Model and Wayne Eckerson’s Analytical Maturity Model.

The Gartner Continuum model was developed based on the difficulty and sophistication of analytics. In Gartner’s model, analytics was divided into four stages: descriptive analytics, diagnostic analytics, predictive analytics, and prescriptive analytics.

• Descriptive Analytics describes the world’s past and present. It answers: What happened? What is happening now? How many, how often, and where? What exactly is the problem? What actions are needed?
• Diagnostic Analytics analyzes historical data to produce insights for future events. It answers: Why did it happen? Why is it happening? What are the trends? What patterns are there?
• Predictive Analytics analyzes past data to predict outcomes of future. It answers: What will happen next? How much sales revenue next year? How will people react to this marketing initiative?
• Prescriptive Analytics answers optimization questions such as: What should I do? How can I make it happen? What happens if we try this? What is the best that can happen?

Wayne Eckerson’s Analytics Maturity Model

While Gartner’s analytics maturity model focuses more on the technical side, Wayne Eckerson tries to combine Davenport and Gartner and provide a comprehensive and holistic approach to achieving analytic maturity. In his book Secrets of Analytical Leaders, Mr. Eckerson introduces an analytics maturity model that helps organizations assess their analytic capabilities and also provides a clear path to analytics maturity. Mr. Eckerson’s model has four axes:

• Analytical Maturity—Analytics evolves from reporting to analysis, then from analysis to the dashboard, and finally from the dashboard to modeling.
• Data Maturity—Data begin from spreadsheets to independent data marts, and then from local warehouses to the enterprise data warehouse, and finally expands to the big data ecosystem that integrates both the internal and the external data.
• Analytics Culture—Firms treat analytics as a cost center in the beginning, then as tactical resources to mission-critical, and finally as strategic resources.
• Scale and Scope—The scale and scope of analytics migrate from individual to departmental, then from departmental to enterprise, and eventually to enterprise plus.

Like Davenport, Wayne Eckerson also drafted a roadmap to analytics success. He developed a very comprehensive analytical framework that highlights the major steps required to run a successful analytical program. Wayne argued that to succeed with analytics, organizations must have the right architecture and data, the right process, the right organization, right people, and the right culture. He cautioned that the ideal path for companies to achieving analytics success was to go from “Flying Blind” straightly up to “Analytic Competitor.” Any other paths would be either too risky or too expensive.

Both Davenport’s and Eckerson’s analytic maturity models are great tools to help organizations recognize their current state on the analytics maturity curve. Not only do these models reveal the critical elements required for the success of analytics, but also plot clear directions for marketers to transform their organizations from analytically impaired to analytical competitors.

Paul Mauschbaugh, Group CIO—Services, Distribution, and Digital at Caterpillar Inc. testified to the accuracy of Davenport’s maturity model. In the article Why Analytics Maturity Matters,6 Paul shared his success story about how he had followed the maturation process and drove analytics into the business at Caterpillar. Paul said he first introduced predictive modeling to the company, but it was not widely accepted by the business side. What he found out was that businesspeople didn’t trust the credibility of the data. Therefore, Paul realized that the first hill to conquer was gaining accountability for the quality of the data within the business community. Paul points out that gaining accountability for the process discipline and thus, the data that are output from that process is a leadership challenge. Establishing KPIs for each line of business and automating the reporting of data against those KPIs is one effective way to drive such accountability. Paul also shared his learning from his amazing journey to analytics maturity at Caterpillar:

• Establish data warehousing/standards to democratize your data to the technical community;
• Gain and leverage data visualization capabilities to democratize your data to the business community; and
• Create and use the right KPIs that measure the results of those business users.

Challenges and Opportunities Facing Small- to Medium-Sized Firms

While these analytics maturity models worked great for Caterpillar, marketers at small- to medium-sized firms are worried that the distance between them and their larger competitors are widening. The reasons for such concern are apparent: First, small organizations do not have the best database, high-quality data, analytic talent, and the analytical tools required to perform analytics work. Second, the biggest problem is money. Small- to medium-sized companies typically have a small budget for improving data and technology, recruiting, and training the best analytics talent. And third, neither Davenport nor Wayne has provided a detailed how-to guide to achieving analytics success for small- and medium-sized organizations. Their models are more a compass than a detailed roadmap.

While it is true that lacking money certainly is a considerable constraint that prevents them from taking advantage of what analytics has to offer, the best-kept secret is that small- to medium-sized firms will benefit more from analytics than their larger competitors. Analytics is an equalizer that democratizes the competitive advantages between large and small organizations. Small- to medium-sized companies have some advantages that their larger competitors do not have. First of all, they usually have fewer business units and organizational layers, which help analytical leaders speed up the decision-making process and gain quicker consensus from different stakeholders, although leaders still have to juggle analytics initiatives with corporate politics. Second, compared to large firms, small- and medium-sized companies are quick and agile, enabling faster implementations of analytics initiatives and faster reactions to risks and learning. Last but not the least, analytics in smaller organizations is more a people and analytics IQ problem than a data and technology issue. Money certainly is an obstacle, but it is not as big a deal as you would expect because there are always workarounds if you are creative enough.

The ALADA Model—Five Pillars of Analytics Success for Small- to Medium-Sized Firms

In general, Davenport’s DELTA model and Wayne’s IQ model can be adapted and utilized by businesses of any size. However, these models were developed mainly for large organizations; therefore, adjustments are necessary when applying them to small- to medium-sized companies. Inspired by Davenport’s DELTA model and based on my successful experience in using data-driven strategies to improve marketing for small- to medium-sized organizations, I developed the ALADA Model—a pragmatic and practical roadmap specially designed for small- to medium-sized companies to achieve analytics success.

The ALADA model has five key success pillars:

• Analytics Marketer;
• Leadership;
• Analytics Scientist;
• Data and Technologies; and
• Analytically Driven Culture.

Difference Between the ALADA model and Other Analytics Maturity Models

Many large organizations started their marketing analytics journey with a focus on data and analysts. Often, the first thing they did was to conquer data accountability and technology upgrades. But for small- to medium-sized companies, while data and technologies are certainly very important, too, the effectiveness and ROI of analytics are largely determined by three types of analytics people—Analytics Marketers, Analytical Leaders, and Analytics Scientists. Together, they determine the analytics IQ and analytics maturity of an organization.

The First Pillar: Analytics Marketers

In most organizations, many dialogues between marketing and analytics take place every day. Most of them, however, are initiated by the marketing side requesting the analytics team to complete specific assignments such as testing product offers or reporting the result of a recently completed direct mail promotion, and so on. In other words, these conversations are centered more at the tactical level rather than the strategic level.

On the marketing side, while marketing does understand that analytics can supply them with significant new decision-making firepower, still, some marketers don’t enjoy talking to analytics because they don’t quite understand the language of analytics. Also, because few marketers have received essential training in statistics and analytics, they might not realize that their internal analytics partners can solve many tough problems if they reach out to their analytical team.

On the analytics side, most analysts are strong in math and coding but don’t have much marketing knowledge and real-world business experience. Due to lack of marketing education and business acumen, some analytics work is not conducted in the context of business. Therefore, the findings and insights while seemingly statistically sound are not actionable, or one insight would cause another or more questions.

Another big challenge for analysts is that most of them are not fluent in business language. Marketers feel that analysts speak a language of their own, not the language of business. According to a 2017 KPMG study, many executives say that they don’t trust their own analytics. It is a shame.

This is when the analytics marketers kick in to help. I borrowed the idea of analytics marketers from Mr. Eckerson. He first defined them as “purple people,” being neither “blue” business leaders or “red” technology leaders but a balanced blend of the two. Because the term “purple people” often demands a further explanation, I came up with a more descriptive name: analytics marketers—individuals who know analytics and also can speak the language of business.

What Does an Analytics Marketer Do?

Analytics marketers are conduits between analytics, IT, and marketing executive decision makers. They translate business requirements into terms that analytics and technologists can understand. Conversely, they also use plain English to show business the value of data and how to exploit it through analytics. Analytics marketers translate insights gleaned from data into actionable recommendations and ultimately into the impact at scale in an organization.

In their role, analytics marketers add value through performing the following tasks within an organization:

• Understand stakeholders’ needs and wants, pain points, and obstacles that prevent them from accomplishing more;
• Assess the quality of data and systems and the functions of current technologies; evaluate the depth and breadth of domain knowledge, analytical techniques, and bandwidth of both the analytics team and the IT support team to identify gaps; and make recommendations and plans to fill these gaps;
• Identify and prioritize business use cases. Analytics marketers leverage their cross-disciplinary knowledge and skillsets to identify the right opportunities that are realistically pursuable. They help business executives identify opportunities that will create the highest value when solved and prioritize the business problems that analytics is suited to solve;
• Collect and prepare data. They help determine the data needed to produce the most useful insights. They work with the analytics team and IT team to improve data quality by cleaning and consolidating disparate data and creating a single customer view that is accessible by stakeholders;
• Help build the analytics engine. They prepare the right tools for performing analytical jobs. They ensure the analytics team will have the right analytics engine that is capable of solving the business problems in the most efficient form for business users.
• Validate insights and make actionable recommendations. They validate and help translate insights into easy-to-understand, actionable recommendations that business users can easily extract and execute.
• Promote data-driven-oriented culture. They are evangelists of analytics. They educate stakeholders to increase the analytics IQ of the company and drive adoptions of a data-driven culture within the organization;
• Help decision makers understand the limits of analytics and how they can go wrong.

A small-sized business needs at least one to two analytics marketers, whereas a medium-sized company will need more. Analytics marketers can be people at any level, but ideally, they are middle- to upper-level managers such as vice president (VP) marketing, VP analytics, marketing managers/directors, analytics managers/directors, or analytics scientists.

The Second Pillar: Leadership

Like any other initiative, support from executives is critical to the success of analytics projects. Davenport considered leadership the deciding DELTA factor. “If we had to choose a single factor to determine how analytical an organization will be, it would be leadership. If leaders get behind analytical initiatives, they are much more likely to bear fruit.”7

Mr. Eckerson also concurred: “The biggest factor that determines analytical success does not involve technology; rather, it involves leadership.”

Broadly speaking, analytics marketers are analytical leaders. However, in the ALADA model, the analytical leader specifically refers to the top executives who sponsor analytics initiatives.

What Does an Analytical Leader Do?

A successful analytical leader does not have to know a lot about analytics. But he or she must recognize that data are critical corporate assets and analytics is the new science of winning. Marketing executive Steve Larkin has helped many retailers to turn around their businesses. I had the pleasure of working with Steve for many years, first at Zales, then Golfsmith, and most recently Charles & Colvard. He always starts his turnaround strategies with fixing the bad customer data and creating a centralized database accessible to both online and offline teams. The improved and consolidated database will lay a good foundation upon which an integrated omnichannel marketing strategy can be implemented. Steve cautions, “It is a needed and often misunderstood and mismanaged topic. Determining and prioritizing the most critically impactful and actionable data is key.”

Essentially, an analytical leader needs to do at least these five things below to ensure successful outcomes of analytics programs:

1. Recruit the right analytics people;
2. Break down company silos to make data accessible and collaboration seamless;
3. Set up rules both written and unwritten to guide how things are to be done in the company;
4. Evangelize the importance of data and analytics to the organization and invest time, money, and people to achieve that vision;
5. Establish a measurement system in which marketing efforts and team performance are regularly evaluated based on actionable KPIs.

Bernie Sensale was my boss when we both worked for Zale Corp almost 20 years ago. Shortly after Bernie became the chief marketing officer of Zale Corp, he quickly identified several major analytics issues with the marketing department:

1. Department silos. Zale Corp once owned six major brands: Zales, Gordon’s, Bailey Banks and Biddle, Piercing Pagoda, and People’s and Mappins Jewelers in Canada. Each brand had its marketing team responsible for doing similar activities such as creating marketing campaigns, working with creative vendors and print workshop, conducting campaign analysis, and so on. The analytical skills of marketing analysts were different across these different brands, but they did not share the best practices among themselves.
2. Data were not updated on time. The organization collected customer and transaction data efficiently, but data were merged, cleaned, and updated by a third-party vendor every 45 days. Lacking the ability to update customer data in a timely manner caused many customer complaints. For instance, a customer requested his or her name be removed from the mail list, but because the call center rep couldn’t update the request immediately, that customer would continue to receive direct mail for at least another 45 days.
3. Campaign results and analysis couldn’t be tracked in real time. An outside company analyzed direct mail and catalog campaigns. They reported the results three weeks after a campaign ended. Learning from previous campaigns couldn’t be quickly applied to the next campaign.
4. Analytics techniques were still at the descriptive and diagnostic stages. The database marketing team’s main job was to perform ad hoc requests such as pulling campaign files using recency, frequency, and monetary, and so on. Their analytic efforts were opportunistic and function specific.

In the next 12 months, I witnessed how Bernie magically transformed an analytically blind organization into an analytical company. What Bernie did was:

1. Hire the right people. Bernie made several important hires. He recruited an analytical director to oversee the analytics programs and a top-level analytics scientist, Dr. Steven Yan, to improve analytical techniques of the whole company.
2. Recognize the outstanding performance of analysts and reward good behaviors. Several top analysts got promoted.
3. Adopt a federated organizational structure to break down the departmental barriers. A central group that consists of the best data scientists and business analysts was created to provide technical guidance and assistance to each brand.
4. Get the data and technologies right. Bernie replaced the legacy campaign management system, purchased more SAS licenses, hired a new data vendor that shortened the data hygiene turnover time, and created a centralized database that was accessible to all brands.
5. Cultivate a learning environment and encourage teams to learn and share the best practices from each other to raise the analytics IQ of the entire company.
6. Evangelize the importance of data and analytics to other executives and middle-level managers.
7. Create a marketing dashboard and corporate reporting package. Campaign results were updated in almost real time. Management used analytics results to help improve decision making.
8. Establish a measurement system across the entire company. Use actionable KPIs such as customer lifetime value and incremental ROI instead of vanity KPIs to measure the effectiveness of marketing initiatives.

Considering Bernie did all these things before Davenport published his breakthrough book Competing on Analytics, even up to date, I am still amazed and very impressed by his textbook-like strategic move.

Matt Corey is the CMO of PGA Tour Superstores. I worked for him during my days at Golfsmith. Matt is a brilliant and charismatic marketing executive. His way to improve marketing results is to hire smart people who have a can-do attitude. Matt’s ability to find so many competent marketers is enviable. Matt understands the importance of marketing analytics. He encouraged the team to embrace advanced analytics and statistical modeling to improve the performance of marketing initiatives. He promoted marketing analytics among senior executives and gained support from IT and merchants. In 2012, SAS voted Golfsmith the best analytics company in the small- to medium-sized company sector.

Those good analytical executives all have similar traits: They are willing to invest in the best people, they take calculated risks and have the courage to challenge the status quo, and they have the ability to gaze into the future.

The Third Pillar: Analytics Scientists

Here an analytics scientist is not an average data scientist or statistician, but one who has:

• analytical intelligence—the ability of abstract thinking, logical reasoning, verbal and mathematical skills, SAS, R, and Python programming skills, and so on;
• creative intelligence—the ability to generate new ideas and deal with novel situations; and
• practical intelligence (also known as “street smarts”)—the ability to apply knowledge to the real world.

Most average analytics professionals have analytical intelligence but are typically weak in creative intelligence, probably even weaker in practical intelligence. A true analytics scientist has a combination of good data knowledge, fluency in programming languages like SAS, R, or Python, mathematical and statistical knowledge, an insatiable curiosity, business acumen, and the ability to creatively solve real-world problems.

The best value an analytics scientist brings to the table is that he or she spearheads new ways to transform current practices. He or she can come up with creative and innovative solutions to solving problems with a very limited budget and technical constraints.

For instance, Golfsmith couldn’t afford to buy a campaign management system, which generally would easily demand a cost up to a quarter-million a year for licensing, training, and a business analyst using the system. So how to solve this issue? Our solution turned out to be quite simple. We worked with the IT team to add a campaign data mart in the database. Then we put the Base SAS on the top of the Oracle database and created an in-house campaign management system that worked far more efficiently than any campaign management system currently available in the market. The invention literally saved Golfsmith more than $250K a year and every year.

The Fourth Pillar: Data and Technologies

Compared to their larger competitors, small- to medium-sized companies typically do not have a centralized data platform that is accessible to all stakeholders. However, being small also has several advantages. First, you can improve the quality of the data with a modest budget. Second, without having to rely on IT support solely, the analytics team can step in and use analytical tools such as SAS, R, or Python to stitch data from disparate databases and create a temp centralized customer database, which can be used by many teams. And third, if you are creative enough, small- to medium-sized companies can create many in-house analytical applications such as the campaign management system that will work equally well as those big-name tools and will save a huge amount of money for the organization. That is why I always say that it is totally possible that small- to medium-sized organizations can achieve the same or even greater significant analytics success with just a fraction of the costs of their large competitors.

The Fifth Pillar: Analytically Oriented Culture

It is difficult to give a concise definition of what an analytically oriented culture is. Basically, an analytically oriented company has the following traits:

• Culture is all about rules and proper procedures. Establishing rules is the most critical step to help cultivate an analytically oriented culture. In an analytically oriented company, people will look at insights gleaned from data before making any decision. Everyone from the chief executive officer to the marketing intern has to follow these rules. For instance, marketing promotion meetings always kick off with facts and lessons learned from history to help shape opinions and strategies for future promotions.
• Data and insights are shared and leveraged across different businesses, organizations, and products.
• An analytically driven company often adopts a federated organizational structure to create a center of excellence to elevate the sophistication of analytics and also allow positive competition among brands.
• An analytically oriented company has many analytics marketers and data translators at different levels. Managers encourage people to ask good questions, challenge assumptions, and have healthy debates on where to make improvements, investments, or other changes.
• People like to challenge the status quo. Innovations in analytics are encouraged and rewarded.
• Actionable metrics and KPIs are used to measure the effectiveness of marketing initiatives and employee performance. Storytelling visual systems that infuse analytics across the organization are widely used.

In a nutshell, the secrets to achieving analytics success are that you must recruit and retain the right analytics people, get the data and technologies right, and create an analytically oriented culture within the organization.

Conclusion

The disappointing results of analytics initiatives were typically caused by one of or a combination of several issues relating to data, technologies, leadership, analytics people, organization structures, process, and corporate cultures. To achieve success with analytics, firms need to take advantage of Davenport’s Five Stages of Analytics Maturity Model to recognize their analytic maturity state and then follow Davenport’s DELTA model and Wayne’s analytical framework to improve their analytics capabilities. Sometimes, a fundamental shift in organizational structure, skills, and culture is required.

Small- to medium-sized companies approach marketing the same way as large companies do. They have unique advantages and disadvantages when it comes to analytics. Complementing Davenport’s DELTA model and Wayne’s analytical framework that is more suitable for large organizations, Mu’s ALADA analytics maturity model was specially developed for small- to medium-sized companies to break through the budget and technical constraints and transform their organizations into analytical competitors as well.

Analytics leaders should also be aware that while these analytics maturity models have revealed the secrets to improving analytics’ capabilities, they are more a compass than a detailed roadmap. Therefore, in real life, innovations in analytics at more granular levels are encouraged and required.

In the rest of the book, we will explore how to leverage marketing analytics to improve ROIs for a variety of marketing channels and how to use the right metrics and KPIs to measure the effectiveness of marketing programs so that marketers can maximize the ROI of their marketing dollars.

---