Customer Intelligence Tames the Big Data Challenge A Harvard Business Review Insight Center Report

Transcription

1 Customer Intelligence Tames the Big Data Challenge A Harvard Business Review Insight Center Report sponsored by

2 2012 Harvard Business School Publishing. All rights reserved. Harvard Business Publishing is an affiliate of Harvard Business School.

3 Customer Intelligence Tames the Big Data Challenge A Harvard Business Review Insight Center Report The HBR Insight Center highlights emerging thinking around today s most important business ideas. In Customer Intelligence, we ll explore how customer intelligence drives innovation, the challenges of having customers on multiple platforms, cutting edge tools and technologies, privacy issues, and more.

4 Table of Contents 1 Mining Big Data to Find New Markets Manish Goyal, Homayoun Hatami, and Angelia Herrin 6 Pin Down Your Customer Intelligence Objectives Tom Davenport 7 Customer Intelligence, Privacy, and the Creepy Factor Larry Downes 9 Marketers Flunk the Big Data Test Patrick Spenner and Anna Bird 10 Tracking the Customer s Journey to Purchase Emma Macdonald, Hugh Wilson, and Umut Konus 11 Turning Customer Intelligence into Innovation Scott Anthony 12 Why Marketing Is King Arthur Middleton Hughes 13 Meet Your Company s New Chief Customer Officer Fatemeh Khatibloo 15 Don t Build a Database of Ruin Paul Ohm 16 Understanding Customers in the Solution Economy David Midgley 17 What Data Can t Tell You About Customers Lara Lee and Daniel Sobol 19 Retailers Turn to Soft Surveillance to Fight Customer Anonymity Robert Plant 20 Customer Experience Should Be Part of Your Business Harley Manning 21 Using Market Research Just for Marketing Is a Missed Opportunity Werner Reinartz 22 Does the 18-to-49 Demographic Matter Anymore? An HBR Management Puzzle Horst Stipp and Jeffrey McCall 24 Use Big Data to Predict Your Customers Behaviors Jeffrey F. Rayport 25 How One Company Uses Customer Data to Drive Sales David K. Williams and Mary Michelle Scott 26 How to Find Out What Customers Will Pay Rafi Mohammed 27 The Big Goal Behind All That Customer Data Bill Lee 28 Free Customers Are More Valuable than Captive Ones Doc Searls 30 Case Study: Should You Listen to the Customer? Thomas J. DeLong and Vineeta Vijayaraghavan iv Customer Intelligence Tames the Big Data Challenge

5 hbr.org webinar summary Mining Big Data to Find New Markets Manish Goyal, Partner, Marketing & Sales McKinsey & Company Homayoun Hatami, Director, Marketing & Sales McKinsey & Company Angelia Herrin (Moderator), Editor for Research and Special Projects, Harvard Business Review OVERVIEW Even though companies today have limited resources, they still desire significant growth. Companies also are able to access more types of data and a greater volume of data than ever before, including real-time data from the Internet and social media. By mining this Big Data, companies can develop insights and identify micromarkets that represent opportunities for growth. When these data-driven insights are translated into specific plans and cascaded to the front lines, companies can produce exceptional results. Tools and technology are important, but even more important are committed leadership, having the right analytical talent, and focusing on specific areas that can create significant value. CONTEXT McKinsey consultants Manish Goyal and Homayoun Hatami coauthors, along with Maryanne Hancock, of the Harvard Business Review article Selling into Micromarkets described how companies can use Big Data to find new markets and drive growth. They provided firsthand examples and responded to numerous questions. KEY LEARNINGS Big Data can produce insights that drive business growth. For years, businesses have had a great deal of data, including CRM data and transactional data. But Big Data is different because of its: Sources Big Data involves pulling together data from multiple internal and external sources. This includes data from customers, channel partners, suppliers, web searches, social media, location data, and even external data such as weather and demographic information. 1

6 Data from large data sets can take the guesswork out of selling. People have always talked about the art of sales. But with Big Data, art is being replaced by scientific analysis. Homayoun Hatami Scale Because so much data is brought together, the amount of data is now far greater than it has been in the past. Timeliness Data is now available in real time to show what customers are talking about in social media (via sentiment analysis) and what they are buying. The value of pulling this data together and analyzing it is to glean new and valuable insights. For example, retailers can use purchase data to estimate a pregnant woman s due date and can target relevant offers. Google used insights from data to modify the font color in its ads, boosting its rate of click-throughs and increasing revenue by $200 million. By mining Big Data, companies can identify micromarkets that represent opportunity for increased revenue. As many companies deal with the reality of constrained resources, they see analyzing and extracting insights from Big Data as a key way to improve the efficiency and effectiveness of their sales and marketing organizations. By putting Big Data at the heart of sales and marketing, insights can be leveraged to improve decision making and innovate a company s sales model, which can involve using data to drive real-time actions. Some of the ways Big Data can be used to drive improved sales and marketing performance include: Identifying micromarkets Big Data, which pulls in external data, can be used to identify new markets, which might be specific customer segments or geographies. Focusing on opportunities Many companies are largely focused on selling to and serving their existing customers. They create goals, budgets, and plans based on looking at past results from current customers. But by using Big Data, organizations can think differently; they can focus not on past results and existing customers but on the opportunities that exist in a market. Big Data will help identify new markets and opportunities that companies might not know even existed. Maximizing sales effectiveness By using Big Data, an organization can assess the targets it is pursuing, the way sales time and resources are being allocated, and what selling propositions and promotions are working best with different customer segments. Prioritizing opportunities rather than existing customers can entail risk. It should be supported by strong analytics. The insights on which decisions are based must be disseminated to all levels of sales and marketing, particularly the front lines. But companies that use analytics have shown that they can utilize their finite resources more effectively to pursue the best opportunities with the right propositions and promotions in the most efficient way. In using Big Data to identify new opportunities, practical prioritization steps include: 1. Defining the right granularity for your micromarkets This might be a specific customer segment, a geography such as a county, or a geographic radius such as 25 miles surrounding a sales rep. 2. Determining the growth potential By using external data, the total opportunity for a micromarket can be estimated with high precision. For example, a maker of plumbing products might look at data such as the number of construction starts in a market and the age of the housing stock. 3. Determining your share of each micromarket With the definition of a micromarket and knowledge of the opportunity, it is possible to determine your company s existing share of a micromarket. 2 Customer Intelligence Tames the Big Data Challenge

7 4. Understanding why there is variance in market share in different micromarkets inevitably, when different micromarkets are identified, a company s market share within them will differ. It is important to understand the underlying reasons for this variance. 5. Prioritizing growth pockets Having used external sources to determine the opportunities that exist, the opportunities can be prioritized and resources allocated to support these opportunities. Case studies showed that organizations generated growth by identifying opportunities and reallocating resources, without adding resources. The Big Data toolkit changes the way that data has traditionally been used. Manish Goyal A key to growth is using Big Data to unlock deeper and deeper insights. An example was shared of a faucet manufacturer analyzing its business, digging ever deeper to glean insights, and using these insights to decide upon actions to grow the business. Insights from this example included finding that: The company s overall market share in California looked strong, which might cause a company to conclude there were no opportunities for growth, but the market share by county varied by 4X. One area with low market share had no sales coverage, and another had no channel partners indicating obvious next steps in both situations. For the geography with no channel partner, analysis of social media identified the most influential partner in that area. In one geography, the company faced a strong local competitor. To compete more effectively, the company s analysis led it to conclude that it needed to better understand customer needs and strengthen its value proposition. It also needed to spend more on marketing to counter competition. Cluster analysis found another US geography with similar characteristics, leading the company to share best practices. The example showed that a granular analysis yielded insights, and each insight led to further analysis and an even deeper insight. As a result of such insights, a company can develop specific actions to drive growth. It is essential to get meaningful insights to the front line. Just developing insights at a headquarters location is not adequate. The speakers recommended building a Big Data toolkit to bring micromarket insights to the front line. It is essential that salespeople understand the insights that have been gained and believe in the validity of the data. This toolkit: Gets data to the front lines so it can be used and acted upon instead of holding data in a central location or in silos. Is predictive and forward-looking, as opposed to looking backward at performance and problems. Is consistent, standard, and repeatable, as opposed to time-consuming and one-off. The key to using Big Data is getting started by focusing on a limited set of outcomes that data can impact. It is important for an organization to realize that using Big Data is a journey. Businesses of any size can embark on and realize value from this journey. A good starting point is to determine a few outcomes that could be improved through analysis and to determine what data is needed to support this analysis. No company will have all of the data it wants, but that shouldn t preclude the need to get started. 3

8 Other Important Points B2B. When analytics in business is discussed, it is often in the context of B2C companies. But this session and the examples discussed showed that analyzing Big Data also has tremendous value for B2B companies in identifying micromarkets and improving sales effectiveness. Leadership. Having leadership that understands and supports the use of Big Data and analytics is an important element of success. Google and Procter & Gamble were mentioned as organizations whose leaders are strong supporters of using data to make better decisions. Organizational structures. Some organizations have a centralized analytics team in strategy, marketing, finance, or another function. With this structure, the focus on sales opportunities often begins when a creative salesperson speaks with a data analyst, who becomes interested in analyzing a business problem/opportunity. In other organizations, analytical capabilities reside in small teams that are distributed throughout the company to support different businesses and functions. They are often closer to the action and might reside in the marketing department or in pre-sales. A key to success is business people working closely with analysts as opposed to handing off a project to them. Emerging markets. There is interest in using analytics to identify and prioritize opportunities in emerging markets. This makes sense because many companies see significant growth opportunities in emerging markets and are resource constrained. A challenge can be a lack of good data from these markets. Learn More For more details about customer intelligence solutions: sas.com/software/customer-intelligence To read more thought-leader views on marketing, visit the SAS Customer Intelligence Knowledge Exchange: sas.com/knowledge-exchange/customer-intelligence To get fresh perspectives on customer analytics from marketing practitioners writing on the SAS Customer Analytics blog: blogs.sas.com/content/customeranalytics To get fresh perspectives on customer analytics from marketing practitioners writing on the SAS Customer Analytics blog: blogs.sas.com/content/customeranalytics Mining Big Data to Find New Markets featuring Manish Goyal and Homayoun Hatami aired on September 5, This webinar summary was created for Harvard Business Review by BullsEye Resources, Inc. The information contained in this summary reflects BullsEye Resources subjective condensed summarization of the applicable conference session. There may be material errors, omissions, or inaccuracies in the reporting of the substance of the session. In no way does BullsEye Resources or Harvard Business Review assume any responsibility for any information provided or any decisions made based upon the information provided in this document. 4 Customer Intelligence Tames the Big Data Challenge

9 SPEAKER BIOGRAPHIES Manish Goyal Partner, Marketing & Sales, McKinsey & Company Manish is a partner in McKinsey s Marketing & Sales Practice. Based in Dallas, he helps clients focus and find growth in micromarkets. Homayoun Hatami Director, Marketing & Sales, McKinsey & Company Homayoun Hatami co-leads McKinsey s Sales & Channel service line and the firm s work in sales growth. He has a broad range of experience working with clients in Europe, the United States, and Asia to power growth through excellence in sales and channels. He is also a leader in knowledge and insight development. Angelia Herrin Editor for Research and Special Projects, Harvard Business Review Angelia Herrin is editor for research and special projects at Harvard Business Review. At Harvard Business Review, Herrin oversaw the relaunch of the management newsletter line and established the conference and virtual seminar division. More recently, she created a new series to deliver customized programs and products to organizations and associations. Prior to coming to Harvard Business Review, Herrin was the vice president for content at womenconnect, a website focused on women business owners and executives. Herrin s journalism experience spans 20 years, primarily with Knight-Ridder newspapers and USA Today. At Knight-Ridder, she covered Congress as well as the 1988 presidential election. At USA Today, she worked as Washington editor, heading the 1996 election coverage. She was awarded the John S. Knight Fellowship in Professional Journalism at Stanford University in

10 Pin Down Your Customer Intelligence Objectives by Tom Davenport that are pretty aggressive on the technology front and those that are more aggressive on the business front. Technologically Aggressive Customer Intelligence Applications Social networks-based offer and attrition models Online-sentiment analysis for social media Targeted next best offers (non-mobile or location-based) A few weeks ago, I was asked to prepare a workshop for a telecom company that wants to invest more in customer intelligence. My first question was, Can we take a week to go through all the possibilities? The problem with customer intelligence is that while everyone wants more of it and better versions of it there are many different avenues to take in pursuing it. No organization can pursue all of them at once (and this company didn t want its workshop to last a week), so the challenge is to narrow featured comment FROM HBR.ORG I hope that this article serves as a reminder and guide to those that read it of the importance of using customer intelligence planning to aid our businesses to succeed. Frank Woodman Jr. down the options fast. You need to reflect on your business situation and goals and consider only the initiatives that will offer the highest impact. For example, customer data and analysis can help you acquire new customers or keep old ones. Depending on which is your current priority, different tools and techniques will apply. The ability to present next best offers to customers (which I and some collaborators wrote about in a recent HBR article) is very high value if you are a retailer with lots of diverse and appealing products to offer, but perhaps not if you are a bank with a set of pretty standard financial services that don t tempt consumers to make impulse buys. Video analytics have great potential for bricks-and-mortar retailers who want to understand their customers, but probably won t do much for an online retailer or a manufacturer who distributes through many channels. Segmentation is a powerful tool, but it only works if you have the ability to treat different customers differently and many consumer products and services companies don t (at least if they are honest with themselves). What I did, therefore, for the telecom team was to draw up a quick list of capabilities an organization might choose to develop in the customer intelligence domain. The notion was that the managers attending would winnow it down for practical purposes at the outset of the workshop, and in the process gain a broad familiarity with the whole landscape of the topic. Perhaps the same list will be useful to you. I divided the possibilities into two categories: those Mobile or location-based offers Non-hypothesis-driven data mining (machine learning, etc.) Uplift or incremental modeling Automated voice of the customer text analysis Automated behavioral targeting for online ad placement Accurate attribution analysis for online advertising Multichannel, multivariate randomized testing Video analytics Business-Aggressive Customer Intelligence Applications Segmentation (and treating different customers differently) Attrition modeling (and taking the necessary actions to avoid it) Predictive models for customerservice episodes Single customer data warehouse with all touchpoints Marketing mix portfolio modeling Adaptive customer profiling (qualitative and quantitative) Loyalty and lifetime value-based pricing Simple or A/B randomized testing By putting something on the technologyfocused list, I meant to underscore that the problems it will cause you to wrestle with are primarily technological and when you successfully implement it, you ll be on the leading edge. On the business-focused list, the primary problems are not technological because the relevant technologies and analytical approaches 6 Customer Intelligence Tames the Big Data Challenge

11 have been around for a while. The primary issues are rather things like getting your entire organization to agree on a common definition of customer, establishing differentiated customer-facing processes for different segments, and so forth. Of course, some of these applications overlap, and some are more infrastructural than others. For example, the single customer data warehouse with all touchpoints would be useful as a building block for most of these applications (but is of course hard to do from a business standpoint). Customer Intelligence, Privacy, and the Creepy Factor by Larry Downes So think of this as the whole toolkit of customer intelligence laid out before you. It s now up to you to be clear on what you re trying to build, and then decide which tool is the best choice for the job. I suppose it would be theoretically possible for someone to create an algorithm or a set of business rules to do this for you to dictate, that is, what customer intelligence applications would make sense for your business. To my knowledge, however, no one has created that. I m curious to hear from HBR readers about how such decisions are being made by smart managers in various settings. What s your biggest problem in customer intelligence, and what does it mean for what applications you need to develop? While we re at it not that this list of 19 isn t long enough already what applications have I missed? u The relationship between large-scale customer intelligence data collection and privacy is more complicated than it seems. From the perspective of data analytics, for example, the bigger the data warehouse the less interesting information about an individual turns out to be. Marketers want to know intimate facts about individual behaviors, but only so they can fit them into increasingly refined demographic groupings of other individuals with similar behaviors and, they hope, similar interests. The more information available about more people, it turns out, the more privacy we actually get as individuals. You really can get lost in a crowd. Perhaps the best explanation for today s resurgent and generalized anxiety about privacy is that it just doesn t seem that way. When a novel information service appears to have zeroed in on one s deepest, darkest secret preferences, it s hard to resist a strong emotional response what might be referred to as the creepy factor. But there is almost always an explanation that, when understood in context, takes the creepiness out of the equation. Gmail users, for example, see ads along the side of the screen advertising products and services that often relate to the contents of recent conversations. We know intellectually that there s no vast army camped out at some Google Ministry of Love, reading through the messages and looking for opportunities to connect them to contextual advertising. It s all software. But that software has gotten so good at reading our minds that it begins to look personal. Only another human and a devious one at that could have connected the dots so easily. That s the moment and we have them more frequently as innovative technologies accelerate their entry into the market when the creepy factor comes into play. Something happens that you didn t expect or hadn t experienced before, and you think, How did they know that? Right now, my Facebook page is showing me photos of three people you may know. I know all three. For two, the connection is obvious. For the third, the connection is eerily indirect. Until I understood what mundane data elements connected all three to me, I felt uneasy about Facebook. The company seemed to be an actual person, and a creepy one at that. As we record more information in digital form in hopes of sharing it with our intimate contacts and, less enthusiastically, with advertisers who pay for the services we love, it s inevitable that more of these visceral responses will occur. When specific data is used in novel ways, the initial response is often the creepy factor. The creepy factor, however, is the response to a novel use of information to provide a seemingly personalized response. Over time, the creepiness decreases. Most of us are now accustomed to customized Google search results, specific Gmail ads, and prescient Facebook recommendations. They no longer make our skin crawl. In response to innovation in customer intelligence, however, privacy advocates are calling for all sorts of new laws to protect us from ourselves. In reality, what they want most is a placebo to cure the creepy factor. Often, there s no need for legislation. Over time, consumers either adjust to what is an 7

12 essentially inert new-information use or act through the market to change the practice. Consumer-enforced change is frequent recent examples include the cancellation of Facebook Beacon and Google Buzz and Apple s modifications to geolocation files stored on consumer devices. When consumers objected to how these services were perceived to be using information, the companies modified their practices or canceled the service altogether. In 2011, to take a specific example, LinkedIn users objected to a new feature called social ads, in which ads for a particular product or service included the profile photos of others in a user s network who recommended it. The creepy factor response was just too overwhelming, and the company quickly agreed to simply list the number of network members who recommended the advertised product. What we ve learned now, wrote Ryan Rolansky, the company s director of product development, is that, even though our members are happy to have their actions, such as recommendations, be viewable by their network as a public action, some of those same members may not be comfortable with the use of their names and photos associated with those actions used in ads served to their network. These are examples where constructive engagement with service providers led to quick resolution true market success. The dangerous alternative to rational discussion is to panic, and in our panic legislate, creating unintended consequences that unnecessarily raise costs for ourselves. After all, the more social the ads at LinkedIn, the more the company can charge its advertisers, keeping subscription fees lower and encouraging a larger and richer network. So a law banning the use of subscriber photos in ads, or something like it, would necessarily raise the cost of a service, perhaps more than the benefit to overall privacy. Legislation should be the last resort, one employed only to correct uses of information that remain disquieting over time. In the US, for example, we have privacy laws that prohibit the use of specific identifying information as a determinant in refusing to transact on reasonable terms. Fair housing, equal employment, insurance red-lining, and sexual harassment laws are all examples. So are, in their own way, privacy laws that restrict the acceptable uses of accurate but often misapplied data elements, including credit histories and health information. These laws in effect correct a creepy factor response that doesn t go away by itself or through market mechanisms. When new applications stimulate our creepy factor response (and more of them will enter the market all the time, thanks to technology trends making data collection and analysis cheaper all the time), the critical policy question becomes what to do during the initial, visceral response period, when creepiness is high. The stakes are high. For better or worse (almost certainly better), Internet users are hooked on the free software, content, and services that rely for revenue on information collection and use. So are the service providers. Legislate too soon and we kill valuable innovation in its infancy. Wait too long and consumers lose faith with the implicit quid pro quo of ad-supported services. My preference is to give the market the first shot. It s faster and cheaper than regulation, less prone to unintended consequences, and easier to tweak after the fact. For those who leap first to legislated solutions to emotional responses, better just to fume, debate, attend conferences, blog, and then calm down before it s too late. In the meantime, more often than not, the creepy factor will go away without the need for intervention. u 8 Customer Intelligence Tames the Big Data Challenge

13 Marketers Flunk the Big Data Test by Patrick Spenner and Anna Bird Today s top-performing marketers as rated by the managers (a profile we call Focusers ) have three key qualities: comfort with ambiguity, ability to ask strategic questions based on data, and narrow focus on higherorder goals. Together, these traits help them filter out noise and apply only the insights or data points that truly matter for longterm success. As marketers get better access to raw numbers and Big Data keeps growing, the importance of this filtering ability will only intensify. The Big Data explosion is driving a shift away from gut-based decision making. Marketing in particular is feeling the pressure to embrace new data-driven customer intelligence capabilities. No wonder a strong appetite for data is one of the most soughtafter qualities in new marketers. And yet, a recent CEB study of nearly 800 marketers at Fortune 1000 companies found the vast majority of marketers still rely too much on intuition while the few who do use data aggressively for the most part do it badly. Here are our key findings: Most rely too much on gut On average, marketers depend on data for just 11% of all customer-related decisions. In fact, when we asked marketers to think about the information they used to make a recent decision, they said that more than half of the information came from their previous experience or their intuition about customers. They put data last on their list trailing conversations with managers and colleagues, expert advice, and one-off customer interactions. But in today s volatile business environment, judgment built from past experience is increasingly unreliable. With consumer behaviors in flux, once valid assumptions (e.g., older consumers don t use Facebook or send text messages ) can quickly become outdated. A majority struggle with statistics When we tested marketers statistical aptitude with five questions ranging from basic to intermediate, almost half (44%) got four or more questions wrong and a mere 6% got all five right. So it didn t surprise us that just 5% of marketers own a statistics text book. Some are dangerously distracted by data While most marketers underuse data, a small fraction (11% in this study) just can t get enough. These data hounds consult dashboards daily and base most decisions on data. They have a plugged in personality type and thrive on external stimulation so they love data and all forms of feedback, including data on marketing effectiveness, input from managers or peers, and frequent interaction with others. We call these marketers Connectors, and they re exactly what most CMOs are looking for. But these types of marketers are actually severe underperformers (they receive much lower performance ratings from their managers than average marketers do). The problem is they don t have the statistical aptitude or judgment required to use data effectively. Every time they see a blip on the dashboard, they adjust and end up changing direction so often that they lose sight of end goals. In management positions, these people can wreak havoc by creating endless fire drills and preventing anyone from sticking with projects long enough to achieve the best results. Worse, many marketing disciplines (especially direct, digital, and loyalty marketing) unwittingly encourage these behaviors and end up magnifying the problem. That s because dashboards often capture response-based metrics such as clicks and open rates that aren t tied to more important measures such as customer loyalty or lifetime value and yet, marketers are rewarded for improving the response metrics. The best focus on goals and filter out noise The bad news for marketing leaders is that the ability to filter out noise is rare (only about 10% of marketers excel here) and hard to teach. The good news is that a wellguided team environment can protect noise chasers from themselves by providing blinkers that keep bright, shiny objects out of view. To drive effective data use, the best marketing leaders reiterate critical business goals constantly (to keep them front of mind despite distractions), teach marketers to put data front and center in their decision making, and sensitize marketers to common data interpretation mistakes. This enables even the most distractible data lovers to overachieve. For more information on the five marketer profiles and the characteristics of today s high-performing marketing teams, download our report here. u featured comment FROM HBR.ORG When you look at the data as individual bits, it can be very overwhelming. Successful marketers are starting to use predictive analytics to bring order and meaning to the data Brian Kardon 9

14 Tracking the Customer s Journey to Purchase by Emma Macdonald, Hugh Wilson, and Umut Konus A customer will touch a company in many different ways before a deal is made. Before you rent your first ZipCar, you ll have talked to friends about it, checked ZipCar s website (and comparison websites), and maybe even called the company. From ZipCar s perspective, all of these touchpoints are important because if you hear bad reports or find the website and call center hard to manage, you ll very likely opt for the safe option of a Hertz or an Avis. Unfortunately, few companies have an overall picture of their customers journey toward a purchase because the information is all too featured comment FROM HBR.ORG To be able to gain insight into the sequence of connections with a brand and the intervals between them and the link to purchasing behavior sounds as significant as segmentation approaches have been up to now. Paula often stuck in a channel silo. An intercept survey that a customer might fill in upon leaving a website can tell you a lot about that customer s experience with the website, but it usually does not provide any information on where the customer will next experience the company. Surveying customers directly after their purchases to explain how they arrived at them means that you have to put a lot of faith in their remembering exactly what they did. A CRM system might let you know how customers moved between the website and the store, but it tells you nothing about how they responded to advertising or word-of-mouth reports. Two years ago, we came across a technique that does allow companies to document quite accurately how their customers actually arrive at a purchase. It is called realtime experience tracking (RET), and we first wrote about it for HBR in a blog last year. It was developed by a market research company called MESH Planning, with which we have been partnering to improve the RET methodology and identify applications for the data it generates. RET involves asking a consumer panel to send text messages on their cell phones every time they come across a given brand or one of its competitors over a period of a week to a month, depending on the length of the purchase process. The structured four-character message captures the brand, the touchpoint type (Saw a tweet about it? Saw it in a shop window?), how positive the customer felt about the encounter, and how persuasive it was. Respondents add further detail online and fill in surveys at the start and end of the study to record brand-attitude changes. Companies can tell how the customer journey works or doesn t from that sequence of text messages. Unilever, for example, could not understand why a campaign for Axe body spray wasn t working in Italy when it was performing well in Poland. In both countries, TV advertising was positively received. But whereas in Poland the ads were followed by high-street touchpoints such as the Axe Police attractive women who would arrest young men and spray them with Axe such reminders close to potential purchase moments were missing in Italy. Traditional econometric models based on spend by media type would have completely failed to pick up this problem. RET can also diagnose how attitudes lead to the next step in the chain, as one major international charity discovered. The charity, which relies on a large network of stores selling both secondhand and new goods to raise both revenue and awareness, recently applied RET in an effort to understand why direct donations (as opposed to store profits) to the charity were falling. The RET project revealed that the in-store experience of customers (and potential donors) was rather mixed; quite a few people felt that the stores were poorly organized and deduced from this that the charity probably wasn t very good at helping its beneficiaries either. They might well purchase goods at the store, therefore, but they did not go on to make donations. Armed with this insight, solving the problem was simple: a smarter layout, displays at the cash register about the charity s fieldwork, and encouraging staff to share their passion for the charity. Non-store donations have since been rising. Because data is gathered in real time, it can be acted on in real time too. PepsiCo recently used RET to fine-tune its relaunch of Gatorade in Mexico, repositioning the brand around sports nutrition. They soon found that experiences in gyms and parks (seeing posters or seeing other people drinking Gatorade, for instance) were twice as effective in shifting brand attitudes as similar encounters elsewhere. They were able to quickly shift more ad and distribution resources into these touchpoints and pass on what they learned as Gatorade was relaunched in other Latin American countries. 10 Customer Intelligence Tames the Big Data Challenge

15 Our first two years working with RET have confirmed its benefits in providing integrated insight, a vital first step toward holistic customer management. Its use is clearly spreading, and doubtless the market research industry will come up with new ways to exploit the rich real-time data RET produces. u Turning Customer Intelligence into Innovation by Scott Anthony It s a paradox of the information age. The glut of information that bombards us daily too frequently obscures true insight. Intelligence should drive better innovation, but unless it is strategically collected and used, it functions like a summer beach novel an engaging distraction. Thoughtful companies intertwine customer intelligence throughout the three phases that characterize most successful innovations. Innovation starts with discovery where an innovator pinpoints an important problem to solve. Ground-level intelligence is critical to this part of the process. While companies are increasingly using detailed analytics to fine-tune pricing, packaging, and product performance, analytics have their limits when it comes to finding the next big idea. After all, data only exists about the past discovering untapped opportunities typically requires a heavy dose of primary research to tease out what the customer needs but cannot easily articulate. Consumers don t do a good job reporting what they currently want or do, let alone what they might want or will do in the future. Procter & Gamble is famous for its deep commitment to these kinds of anthropological approaches. For example, in the early 2000s, P&G investigated the cleaning habits of Indian consumers who washed garments by hand. At first glance, that s a counterintuitive place to look for new growth, because those consumers are unlikely to buy P&G detergents formulated for washing machines. But since hand-washers constitute 80% of the home-based washing market in India, it was too big a market to ignore. P&G observed that many consumers were in fact hand-washing garments using machine-oriented detergents to take advantage of their superior cleaning benefits. However, the chemical formulations weren t intended for hand-washing and could cause abrasions or burns. Insight in hand, innovators next blueprint a solution to address the identified problem. In the case of P&G in India, the idea the company ultimately commercialized was Tide Naturals, a special formulation that lets hand-washers get the cleaning benefit of Tide without suffering the downsides of machine detergents. There are substantial opportunities to generate real-time intelligence by involving customers in the blueprinting process. For example, four years ago, the Indian company Godrej & Boyce was working on an idea for a small, battery-powered refrigerator to reach the 80% of Indians without refrigerators. The team working on the idea brought an early prototype of the concept to a rural village and showed it to 600 women. Navroze Godrej, who leads the company s disruptive growth efforts, describes how the event was a way to get instant feedback, allowing Godrej to co-create with these women. It was also here that the final color we went with ruby red was decided pretty unanimously with 600 women. P&G has a number of mechanisms to facilitate this kind of real-time customer input, such as a specially designed Home of the Future and Store of the Future 30 miles north of corporate headquarters in Cincinnati; online networks such as VocalPoint, where hundreds of thousands of mothers provide feedback on products; and the regular practice of bringing real consumers into its offices. The final stage in the process is to iteratively test an idea by executing smart experiments to test key assumptions. Does the product (or service) solve the consumer problem it was intended to in a way that generates repeat use and repurchase? Will consumers purchase at the required price point? Can the idea be reliably delivered at scale? Do the economics work? Ideally, tests to answer these kinds of questions aren t run in the laboratory, but with real customers in everyday settings. Presenting early ideas to customers to get their feedback provides vital intelligence that helps increase the success rate and sustainability of innovation. featured comment FROM HBR.ORG This is a good reminder that customer data serves multiple purposes and that prospects are often the best sources of new product ideas. Don Nanneman 11

16 There are a number of ways to generate this kind of real-world input. Many consumer-facing companies use employees as customers to test new concepts inside their walls. For example, the headquarters of Unilever s India operations contains a street with shops and kiosks selling Unilever products to glean insights from employee customers. Business-to-business companies can consider bringing rough ideas to customer councils, running pilots with select customers, or even using booths in industry conferences to gauge interest in new ideas. Since the goal is learning, companies should ensure that they keep tests simple and focused. Affordable online tools such as LinkedIn, elance.com, Survey- Monkey, Wix, Amazon s Mechanical Turk, Appmkr.com, and Google SketchUp can complement live testing to accelerate effective, affordable experimentation. Companies seeking to more formally intertwine intelligence with innovation should consider three straightforward starting points: 1. Mandate that everyone in the company increase the amount of time they spend with customers however much time your company is spending, it is probably not enough. 2. Find simple ways to make customer conversations more frequent. Consider forming a lead user panel or creating an online community such as those offered by CommuniSpace. 3. Build a little-bets lab, a mechanism by which you can selectively introduce early ideas to the market. For example, at beta620.nytimes.com, users can test-drive early experiments offered by The New York Times Company. Littlebets labs facilitate the thoughtful process of strategic experimentation that typifies successful innovation. Want more intelligent innovation? Start by intertwining intelligence and innovation. u print, or direct mail ad is what it is. On , the ad is much more. Because of electronic links, those who open your s can do their own research: they can explore and see any of the thousands of products you sell. They can see the colors and sizes. They can, and they do, read ratings and reviews. They can put products in their shopping carts and buy them. Fine, say the TV folks, but shopping cart sales through s are seldom more than 5% of total sales. Nothing to write home about. What these detractors seem to willfully ignore is that s create impressions that lead to sales through other routes. Some of these routes can be tracked. The recipient can open it or delete it. If she opens it, she can click on it, perhaps buy something or print out a coupon and take it to a store. Finally, if she puts things in her cart but does not buy, you can send her an abandoned shopping cart that usually yields 29% of lost sales. Why Marketing Is King by Arthur Middleton Hughes In a business world obsessed with gaining more customer intelligence, you would think that marketing would get more respect. But just look at media spending. According to emarketer, this year US companies are spending about $64 billion on TV, $34 billion on print ads, and $39 billion on Internet advertising. And how much are they are spending on ? For that, we have Forrester data: only about $1.5 billion. Of course, compared to other media, messages are dirt cheap to send. With TV, you are spending on ad agencies, creative studios, and cable channels. With print ads, you are helping to keep newspapers and magazines alive. Direct mail costs more than $600 per thousand pieces. With , there are almost no costs at all. But its low cost only makes the argument stronger that marketing is the most cost-effective advertising method available today. Certainly beats the competition from a measurability standpoint. With TV, you do not know who is watching your ads. Ditto with print. Even with direct mail, you cannot be sure that your mail has been delivered or that anyone reads it when it gets there. With , you know within 24 hours exactly which messages have been opened, by whom, what links the openers clicked on, and what part of your message was working. A properly structured message provides this benefit to the marketer because it provides benefits to consumers. A TV, But note that, in many cases, she also does things that are hard to track. She can get in her car and drive to a mall to buy the product. She can pick up her phone and order it. She may be prompted to do research on Google for better prices of similar products or discuss the offer with her spouse or a friend, leading to a possible purchase later. These are all the behaviors that provide the rationale for TV or print advertising. My point is that s prompt the same kinds of behaviors. Thus, there is an off multiplier. For every purchase in an shopping cart, we can fairly assume that there are some number of other nontracked profitable purchases that occur because of the arrival of the a number that quantifies all the non-tracked behaviors that recipients engage in. If you are going to make a case for investing more heavily in marketing, you have to determine this off- multiplier to account for all the sales your s can be expected to generate. How can that be done? A retailer I ve worked with that has 900 stores and is very active with campaigns recently did a great study. It took a group of 105,000 customers in its loyalty club database, divided them into three groups of 35,000, and marketed to 12 Customer Intelligence Tames the Big Data Challenge

17 What are the skills that enable a CI leader to take on the formidable challenge of owning the customer relationship and being the organization s customer advocate? There are three: the three groups differently, as shown in the chart below (click to see a larger version). Thanks to the loyalty program, it was able to see all subsequent purchases by these customers. Direct mail has a higher response rate than . But note that direct mail costs about 100 times as much. Meanwhile, the data collected by the retailer allowed it to calculate its off- multiplier (a simple matter of dividing the percentage of online sales by the percentage of in-store sales generated by -only marketing). It is In other words, for every shopping cart sale, this retailer gets 3.76 other, typically non-tracked sales due to the . What might your off- multiplier be? Zero is of course possible, but studies to date suggest that a number between two and three is typical. Once you factor in your off- multiplier, it s a very safe bet that will beat all your other marketing methods in terms of return on investment. As marketing gains more respect, marketing intelligence will meet customer intelligence. u Meet Your Company s New Chief Customer Officer by Fatemeh Khatibloo Customer intelligence is at an organizational inflection point. This practice, which is largely the evolution of database marketing, has become a critical driver of business strategy for global organizations in nearly every industry and vertical because it supports decisions with data. In this way, CI s value extends well beyond the marketing organization. But what does a successful CI professional s future look like? The answer lies in the rise of a new type of executive: the Chief Customer Officer. Why CI pros should aim for the CCO job A decade ago, the only option for an ambitious CI professional was the CMO s office, even though in most enterprises, that seat will be filled by an executive with brand and media expertise. Today, the new Chief Customer Officer role a position that I believe most smart companies will create in the upcoming decade is well-suited to the skills of a CI leader. My colleague, Paul Hagen, writes extensively about the emerging CCO and was good enough to share some of his raw data with me. In his research, he reports that these individuals typically have extensive sales, marketing, and operational backgrounds. But when we dug more deeply into their previous work history, we found titles such as Consumer Engagement Lead, VP, CRM & Loyalty Program Development, and VP, Direct Response Marketing all customer intelligence positions. The ability to interrogate data. I m not suggesting that these folks need to run SPSS or sit in front of a business intelligence tool all day. But they do understand the importance of data in addressing business questions. They know how to build a hypothesis, mine the data, test a solution, and validate. And most importantly, these individuals can translate data insights into strategic business language to gain adoption and credibility for their approaches. The ability to speak IT. Marketing and IT have never been more interconnected. But there s still a fundamental language disconnect between these groups. Many customer intelligence leaders have bridged that gap already they know how to translate data insights into the language of business to gain credibility and drive adoption. As organizations transition from product-centric to customer-centric, CI leaders are ideally positioned to build a data-driven business case to justify the organization s marketing technology and business intelligence needs. The ability to describe customers realistically and actionably. For decades, agencies and market researchers described customers in terms of a handful of archetypes and personas. But that s not realistic in today s omni-channel, multidevice ecosystem because marketing s promises of right place, right time, right offer require microsegmentation and granular customer understanding. CI leaders already understand this. They manage preference centers, they use propensity models to define and meet customer needs, and they leverage tools to recognize customers across channels and devices. The future of customer management means individually optimized touchpoints, and CI pros are already halfway to the finish line. What skills should CI leaders hone to succeed as CCOs? The role of the Chief Customer Officer will vary by industry and organization. But 13

18 Hagen has identified three attributes that these leaders must possess: 1) a passion for customer experience, 2) a strong personal brand, and 3) operational know-how. These leaders must: Act as the voice of the customer. It sounds like a platitude. But to be recognized as an organization s customer advocate, CI pros should use their access to data to create customer journey maps that track customers across all channels and touchpoints and introduce these tools across the organization at every opportunity. Use the maps to question and challenge decisions that impact the customer and support these challenges with data and insights. featured comment FROM HBR.ORG This is a brilliant piece! As a Business Development Specialist for a major Fortune 50 company, we ve been moving in this direction for some time now. Jeremiah LeBlanc Become tempered radicals. In October 2001, HBR published a piece by Debra Meyerson in which she advocated four approaches to becoming a unique type of change agent: 1) disruptive selfexpression, 2) verbal jujitsu, 3) variableterm opportunism, and 4) strategic alliance building. Leaders who practice this type of change management gain trust and build a personal brand that earns respect without strong-arming. CI leaders, in particular, should learn how and when to apply these methods because without them, it can be hard to gain executive-level visibility. Get operational expertise, and fast. CI leaders often spend time delivering insight to colleagues in operational functions. But instead of being just a service provider, engage these individuals at a more strategic level. Consider trailing field sales agents or following a product development cycle from inception to market. Consider how these organizations and processes use data, how they function, and what operational challenges they face. Lead cross-functional task forces outside of the normal marketing and IT purviews, and look for opportunities to pilot and lead entire customer-focused functional teams. I believe that customer intelligence leaders, with their deep and broad understanding of customers, are the natural choice to lead organizations along the path to true customer centricity. They are the future CCOs. Agree? Disagree? I d love to hear your thoughts. u 14 Customer Intelligence Tames the Big Data Challenge

19 Don t Build a Database of Ruin by Fatemeh Khatibloo Many businesses today find themselves locked in an arms race with competitors to see who can convert customer secrets into the most pennies. To try to win, they are building perfect digital dossiers to use a phrase coined by Daniel Solove massive data stores containing hundreds, if not thousands or tens of thousands, of facts about every member of our society. In my work, I ve argued that these databases will grow to connect every individual to at least one closely guarded secret. This might be a secret about a medical condition, family history, or personal preference. It is a secret that, if revealed, would cause more than embarrassment or shame; it would lead to serious, concrete, devastating harm. And these companies are combining their data stores, which will give rise to a single, massive database. I call this the Database of Ruin. Once we have created this database, it is unlikely we will ever be able to tear it apart. I have become convinced that my earlier bleak predictions about the Database of Ruin were in fact understated, arriving before it was clear how Big Data would accelerate the problem. Consider the most famous recent example of Big Data s utility in invading personal privacy: Target s analytics team can determine which shoppers are pregnant, and even predict their delivery dates, by detecting subtle shifts in purchasing habits. This is only one of countless similarly invasive Big Data efforts being pursued. In the absence of intervention, companies will soon know things about us that we do not even know about ourselves. This is the exciting possibility of Big Data, but for privacy, it is a recipe for disaster. If we stick to our current path, the Database of Ruin will become an inevitable fixture of our future landscape, one that will be littered with lives ruined by the exploitation of data assembled for profit. But we can chart a different course in various ways. I think our brightest engineers can develop innovative privacy-enhancing technologies that will enable new techniques for data analytics that minimize costs to privacy. I hope that public institutions and industry, through self-regulation, will devise ways to better balance the burdens on privacy and the benefits of Big Data. If nothing else, I anticipate that society will slowly develop new norms for engaging with the massive amount of information collected about us, creating informal rules governing when and how it is appropriate to release, collect, and use data, the way minors have learned to speak and listen carefully on social networks. But every one of these correctives requires the same thing: time. We need to slow things down to give our institutions, individuals, and processes the time they need to find new and better solutions. The only way we will buy this time is if companies learn to say no to some of the privacy-invading innovations they re pursuing. Executives should require those who work for them to justify new invasions of privacy against a heavy burden, weighing them against not only the financial upside but also against the potential costs to individuals, society, and the firm s reputation. Companies should do this not only as a matter of good corporate social responsibility but also because it will likely square with the government s recommendations for protecting privacy, which seem to advise caution and deliberation, under the banner of context. Earlier this year, Federal government officials released two privacy reports the White House s white paper and the FTC s final privacy report that together describe a national privacy policy for the foreseeable future. Although the two reports vary on some particulars, they both point to context as a central, important, and fundamental measuring stick we should use to assess decisions that bear on personal privacy. The FTC report offers three broad recommendations: Privacy by Design, Simplified Choice for Businesses and Consumers, and Greater Transparency. In discussing the second recommendation a call for simplified and more transparent choice the FTC suggests a carve-out. Companies do not need to provide choice before collecting and using consumer data for practices that are consistent with the context of the transaction or the company s relationship with the consumer, or are required or specifically authorized by law. Under this standard, it might be consistent with the context, for a company in a direct business relationship with a customer to use that customer s information to deliver ads for its other services, but it might be inconsistent with the context thus requiring notice and choice to sell that information to third-party advertisers, the FTC explains. Similarly, the White House white paper defines a Consumer Privacy Bill of Rights, which would protect, among other things, Respect for Context. Consumers have a right to expect that companies will collect, use, and disclose personal data in ways that are consistent with the context in which consumers provide the data, the paper explains. These parallel pronouncements mean that companies that deal with personal information (meaning all companies, really) need to focus much more often than they featured comment FROM HBR.ORG Paul, I ll be quoting your piece in tomorrow s Forbes online. Thanks for the interesting ideas. Chris Taylor 15

20 have on the history of privacy practices in their industries. Although neither report defines in depth what it means by the word context, to me the message seems to be do not push the privacy envelope. Companies that use personal information in ways that go well beyond the practices of their competitors risk crossing the line from responsible steward to reckless abuser of consumer privacy. The lesson is plain: compete vigorously and beat your competitors in every legitimate way, except when it comes to privacy invasion. Too many companies have learned this lesson the hard way, launching invasive new services that have triggered class action lawsuits, Congressional inquiries, and media firestorms. These companies knew that they were treading where others had feared to go. This may have felt like an exciting opportunity. It should have felt instead like perilous risk-taking because it meant hurtling beyond the contextual borderlands defined by past practice. u Understanding Customers in the Solution Economy by David Midgley Companies in all varieties of B2B markets have moved beyond selling products and services to offering complete solutions to their customers. Alstom keeps trains ready to run each morning for railroad operators rather than just selling the rolling stock to them. General Electric helps hospitals manage and use patient data rather than selling them the equipment and software to do the job. Hilti provides and maintains power tools for builders. Rolls-Royce runs the engines you see on the wings of your plane. Syngenta offers rice farmers planted fields. From the provider s perspective, selling solutions allows companies to differentiate themselves in commoditizing markets and to benefit from economies of scope across multiple profit and service capabilities. For customers, these solutions offer better value than the products and services that went before. After all, who would not prefer a solution to their business problems rather than simply buying services and products? My INSEAD colleague, Professor Markus Christen, and I have been researching solution strategies in a number of industries. We believe that it is the way of the future and that we are moving toward a solution economy where organizations focus on what they are really good at, relying on their suppliers solutions to take care of the rest. Getting there, however, is going to involve a bigger change than most suppliers realize most particularly in the way they gather information about their customers. For a start, there has to be general agreement on what the word solution actually means. Customers and suppliers often have different definitions. B2B customers regard a solution as something that helps their business. That is, a solution increases their revenues, lowers their costs, or reduces their risks and in doing so boosts their overall profitability. The trouble is suppliers don t always think about their solutions from that perspective. Many define a solution as a package or bundle of the products and services they already offer. And what they already offer may have no explicit link to an individual customer s business objectives, since the bundles of products and services are constructed to meet generic needs. Achieving a solution economy will require these providers to change their mindset. But this is not the whole answer. The solution itself has to add real value not seen before. If not, for all the interest shown by the supplier, the customer is going to view the solution simply as a volume discount offer. Creating that new value will require suppliers to combine their expertise with their understanding of the customer s business needs. This calls for changes in how B2B companies gather customer intelligence. Specifically, customer researchers need to: Ask different questions much more often. Most companies customer research focuses on how the customer uses or perceives the supplier s products and services rather than on how these help achieve the customer s business objectives. There needs to be a change, therefore, in the types of questions companies ask their customers and how the resulting data is interpreted. What s more, traditional market research is episodic, whereas the nature of solutions requires a more ongoing, relationshipbased approach. Periodic market research surveys or focus groups should be replaced by continual data collection from all interactions between the supplier and its customers. Observe the customer directly. Solutions often represent major innovations and, as such, customers may not always be able to grasp the value immediately. Traditional research techniques (surveys, focus groups, etc.) need to be complemented with observation of how the customer currently operates. And not observation by typical market research or salespeople rather, observation by personnel with the necessary technical and business expertise to spot opportunities. This is, of course, a challenge when there are large numbers of customers. Technology can help here. One company we know maintains sensors in customers factories so that it can monitor how the relevant parts of the customers operations work. The bottom line is that customer intelligence for the solution economy will look 16 Customer Intelligence Tames the Big Data Challenge