Can Big Data Protect A Firm From Competition?

5m ago
234.95 KB
20 Pages

Can Big Data Protect a Firm from Competition?Anja Lambrecht and Catherine E. Tucker December 18, 2015 Anja Lambrecht is an Assistant Professor at London Business School, London NW1 4SA, UK. Catherine Tuckeris the Distinguished Professor of Management Science at MIT Sloan School of Management, MIT, Cambridge, MA02139, USA. The authors thank the Computer and Communications Industry Association for generous funding ofthis research. All mistakes are our own.1Electronic copy available at: 2705530

Contents1 Introduction42 Is Big Data Inimitable?53 Is Big Data Rare?64 Is Big Data Valuable?85 Is Big Data Non-Substitutable?116 Implications152Electronic copy available at: 2705530

Executive SummaryThere is plenty of hype around big data, but does it simply offer operational advantages, orcan it provide firms with sustainable competitive advantage? To answer this question, we lookat big data using a classic framework called the ‘resource-based view of the firm,’ which statesthat, for big data to provide competitive advantage, it has to be inimitable, rare, valuable, andnon-substitutable.Our analysis suggests that big data is not inimitable or rare, that substitutes exist, andthat by itself big data is unlikely to be valuable. There are many alternative sources of dataavailable to firms, reflecting the extent to which customers leave multiple digital footprints onthe internet. In order to extract value from big data, firms need to have the right managerialtoolkit. The history of the digital economy offers many examples, like Airbnb, Uber and Tinder,where a simple insight into customer needs allowed entry into markets where incumbents alreadyhad access to big data.Therefore, to build sustainable competitive advantage in the new data-rich environment,rather than simply amassing big data, firms need to focus on developing both the tools andorganizational competence to allow them to use big data to provide value to consumers inpreviously impossible ways.3

1IntroductionThe digitization of the offline and online economy alike means that firms are naturally collecting’big data’, distinguished by its volume 1 , variety of formats spanning text, image and video, andvelocity, meaning that data is recorded in real time.2There is plenty of hype around big data. Firms are constantly exhorted to set strategies inplace to collect and analyze big data (Bughin et al., 2010; Biesdorf et al., 2013), and warned aboutthe potential negative consequences of not doing so. For example, the Wall Street Journal recentlysuggested that companies sit on a treasure trove of customer data but for the most part do notknow how to use it.3 Recent articles such as McGuire et al. (2012) and McAfee et al. (2012) havemade cases for why big data offers a short-term operational advantage, both in terms of cost andperformance, for firms who find ways of using it successfully.However, big data’s long-term strategic, rather than operational, implications for firms are lessclear. Academic opinion differs on whether it will lead to a new type of competitive advantage(McGuire et al., 2012) or not.4 The question of whether big data can indeed confer sustainablecompetitive advantage is critical for firms but has, to our knowledge, received surprisingly littlesystematic attention.To evaluate the strategic role of big data as a source of sustainable competitive advantage oras a barrier to entry, we use a classic framework in strategic management sometimes referred to asthe ‘resource-based view of the firm’ (Wernerfelt, 1984; Barney, 1991; Peteraf, 1993; Barney, 2001).This literature is useful because it sharply distinguishes factors that enhance an entire industryfrom a ‘sustainable competitive advantage’ that benefits a single firm. For there to be a sustainablecompetitive advantage, the firm’s rivals must be unable realistically to duplicate the benefits of1Companies such as Amazon and Walmart already work with petabytes of data in a single data set (McAfee et al.,2012).2This functional definition of bid data does not specify the depth of consumer insight it can provide. Big dataspans anonymized user data, personally identifiable information, search query data, web browsing data or data onconsumer sentiments or purchase intentions. Depending on the specific type of data under consideration, how valuableit is to the firm may alue-of-customer-data-1444734633?mod djem jiewr digital-strategy-works-and-yours-probably-doesnt. This article highlights that because digital technologies are visible and accessible to competitors, it is hard to generate acompetitive advantage.4

this strategy or input. Specifically, Barney (1991) suggests that for a firm resource to be a sourceof competitive advantage, the resource has to be inimitable, rare, valuable, and non-substitutable.In a similar spirit to Markman et al. (2004)’s analysis of patents, we examine along each of thesedimensions whether big data is a source of sustainable competitive advantage to firms.2Is Big Data Inimitable?For big data to be inimitable, no other firm should easily be able to replicate the advantage. Thereare two underlying economic reasons for why big data in many instances is unlikely to be inimitable.First, big data is non-rivalrous, meaning consumption of the good does not decrease its availabilityto others. Second, big data has near-zero marginal cost of production and distribution even overlong distances (Shapiro and Varian, 1999). These two basic characteristics, combined with the factthat customers constantly leave footprints on the internet, have lead to a thriving industry whereconsumer big data is resold.This type of commercially available big data allows new entrants to gain insights similar to thoseavailable to firms that own big data on a large number of customers. There are many examples oflarge commercially available data sets. Acxiom has ‘multi-sourced insight into approximately 700million consumers worldwide’ with over 1,600 pieces of separate data on each consumer; Datalogixasserts that its data ‘includes almost every U.S. household.’5 Comcast is planning to license TVviewing data collected through set-top boxes and apps.6 Other companies, such as the Oracleowned Bluekai, sell cookie-based user information online to allow for targeting advertising basedon a user’s past activities or demographics. Bluekai states that it has data on ‘750 million uniqueusers per month with an average of 10-15 attributes per user.’7 To protect both their customersand themselves, such companies undertake to ensure that their data collection complies fully withdata protection rules.5See Acxiom Corp., 2013 10K Annual Report for the Period Ending March 31, 2013 and Staff of S. Comm. onCommerce, Sci., and Transp., Office of Oversight & Investigation, A Review of the Data Broker Industry: Collection,Use and Sale of Consumer Data for Marketing htm#DSMKT36165

Given the different possible types of big data, an obvious question is whether this analysisextends to cases where the big data has what appears to be unique or individual insights. Forexample, recently the retail store Target hit the headlines because of its alleged ability to use itsretail shopping data to predict a pregnancy even before close relatives knew about it.8 However,even such highly specific and timely data-driven insights are easy to imitate for firms that do notown a national database of retail sales. For example, a marketing unit of the credit-scoring agencyExperian sells frequently updated data on expecting parents, along with income and first-birthinformation.9In addition, data that is available due to individual consumer-level tracking is complementedby the explosion of user-generated content where consumers themselves create a footprint of theirbehavior, likes, opinions and interests across the internet. Recent research in computer science hasemphasized that by combining a myriad of external online profiles, external firms can gain hugeinsights into any one customer (Narayanan and Shmatikov, 2008; Calandrino et al., 2011). Firmscan also use such content as a direct substitute for customer data. For example, Edelman (2015)discusses that was able to build a successful home-buying digital platform by relyingon existing town assessment data.10In short, where a market for data exists, it is unlikely that big data is inimitable.3Is Big Data Rare?For Big Data to be a ‘rare’ resource would mean that few other firms possess it. However, there aretwo reasons why this is unlikely to hold. First, large shifts in supply infrastructure have renderedthe tools for gathering ‘big data’ commonplace (Greenstein et al., 2013). Cloud-based resourcessuch as Amazon, Microsoft, and Rackspace make these tools not dependent on scale11 and /shopping-habits.html? r 0 There are some doubts over theorigin of this story and whether Target actually did this - see for example .com/small-business/prenatal-lists.jsp10He highlights the interaction between publicly available information and user generated content, saying, ‘Zillow’s initial information was good enough to attract consumer interest, at which point property owners happilycontributed corrections, photos, and other information. Indeed, real estate agents were soon willing to pay to showtheir advertisements in and around Zillow’s property ring-the-top-three-cloud-storage-providers/6

costs for data continue to fall, so that some speculate they may eventually approach zero.12 Thisallows ever smaller firms to have access to powerful and inexpensive computing resources. Free opensource technologies such as Hadoop that allow users to analyze large datasets are widely availableand accessible.Second, as consumers’ lives increasingly shift to the web, consumers leave traces of their needsand preferences everywhere. Firms who embrace these low-cost digital technologies have manyopportunities to gather customer data. Telecom companies can collect data on calling behavior andbrowsing on their phones; Amazon, Macy’s and Walmart collect detailed consumer-level purchasedata, while platforms such as Bluekai collect a large range of detailed consumer browsing andpurchasing information across multiple website.13Indeed, such ‘multi-homing’, that is the use of multiple different digital services by consumers,means that similar pieces of information are often available to many different companies. Take, asan example, consumers who use multiple online social media such as Facebook, Twitter, LinkedInor Instagram and share broadly similar information through each of them. Or, consider access toinformation in the app ecosystem: Many apps, and not only those related to location or weather,regularly ping location data - as many as hundreds of times a week - meaning that a user’s locationis always available to a wide range of firms (Almuhimedi et al., 2015). Of course, as we later discuss,these firms still have to invest in ensuring that they have the technical skills to transform this datainto valuable insights.Seeing that big data is not inimitable or rare, we turn to the question of whether and when bigdata is valuable for o1.html13The European Commission spoke similarly in 2014 when concluding its investigation into Facebook’s acquisitionof WhatsApp. It concluded that ‘there are currently a significant number of market participants that collect user dataalongside Facebook, including Google, Apple, Amazon, eBay, Microsoft, AOL, Yahoo, Twitter, IAC, LinkedIn, Adobeand Yelp and that, in addition, ‘there will continue to be a large amount of Internet user data that are valuable foradvertising purposes and that are not within Facebook’s exclusive control’. See (Tucker and Wellford, 2014) as wellas ”Case No COMP/M.7217 - FACEBOOK/ WHATSAPP”, sions/m7217 20141003 20310 3962132 EN.pdf7

4Is Big Data Valuable?Much of the current managerial literature is focused on whether or not ‘big data’ is indeed valuablefor firms in that it enhances a firm’s ability to have profitable relationships with customers (Chenet al., 2012). Cuzzocrea et al. (2011) point to three open problems currently challenging analystsand researchers faced with ensuring that big data is valuable to organizations. We discuss thesechallenges in turn and conclude that by itself big data is not sufficient to create profit-enhancingopportunities.The first challenge limiting the value of big data to firms is compatibility and integration. One ofthe key characteristics of big data is that it comes from a ‘variety’ of sources. However, if this data isnot naturally congruent or easy to integrate, the variety of sources can make it difficult for firms toindeed save cost or create value for customers. Such hindrances may prove particularly burdensomein industries such as healthcare, where prior research has shown that firms have strategic incentivesto ensure that data is siloed and hard to integrate (Miller and Tucker, 2014).The second challenge to making big data valuable is its unstructured nature. As discussed byFeldman and Sanger (2007), specialized advances are being made in mining text-based data, wherecontext and technique can lead to insights similar to that of structured data, but other forms of datasuch as video data are still not easily analyzed. One example is that, despite state-of-the-a