[Music] how can we design strategy to make data valuable and create a competitive advantage this is the question we explore today in my video on the machine learning value chain we have established that data is part of a larger system and discussed whether more and or better data can help firms to make better decisions if you haven't watched this video i suggest that you do this first and then come back here let's see how we can think about data strategy we will cover the different ways to gain a competitive advantage from data in addition we
will take a dynamic perspective and describe the transformation process that enables firms to become data driven to put all this into perspective we will talk a bit about how facebook uses data strategically in a way you might not be aware of so how can we gain a competitive advantage from data let's start with a little reminder regarding the different ways scholars in strategic management have characterized where competitive advantage can come from first is the idea that performance in an industry is mostly defined by the characteristics of the industry itself hence it is important to take
market conditions into account when developing strategy for the firm the relational view puts the networks and relations between internal and external stakeholders into focus organizing these relationships is key for achieving a competitive advantage and finally there is the idea that it is the stock of internal resources and capabilities that define why some firms perform better than their competitors this is the resource-based view and the perspective we want to focus on today we can say that a firm can gain a competitive advantage if it has resources that are valuable rare and costly to imitate and if
the firm is organized such that it can capture the value of the resources in short v r i o so far so good now let's try to add data to this framework there are four questions that we need to ask the first one is does data enable our organization to obtain valuable insights to exploit new business opportunities and or neutralize competitive threat so is the data actually valuable the second question is in regard to the rarity of the data is our data rare can a few of our competitors acquire or process them the third question
asks whether it is costly to imitate do our competitors without data face challenges or obstacles in obtaining or developing it is it difficult for our competitors to imitate what we can do with our data the final question is does our business strategy our way to organize and our corporate culture support the exploitation of valuable rare and costly to imitate data resources this is where the machine learning value chain comes to mind again right to benefit from data as a valuable and perhaps rare resource embedding it in the broader value chain is important only in combination
with models and algorithms and computing power it becomes costly to imitate and therefore can create a competitive advantage time for an example consider facebook and their data strategy facebook of course has lots of data on the preferences of their customers users literally tell facebook what they like but that's only true for pictures stories and products that these users see on their facebook timelines facebook knows what you like from observing your behavior within their own platform this is information that facebook's competitors don't have it's rare and not easy to imitate however facebook's competitors like in a
broad sense also have data that is rare and potentially valuable there are many things on the internet that you may like or dislike and facebook collects data about that by convincing website owners to place buttons that say like or share on their sites every time you load such a button or click on it facebook is collecting data on your browsing activities outside of facebook's platform how is this type of data valuable to facebook in a press release from september 2015 facebook says we sometimes hear from people that the ads that they see are not as
useful or relevant to them as they could be and therefore we are solving this problem by introducing online interest-based advertising of course not every organization is as data savvy as facebook for many organization thinking about the value of machine learning and data is pretty new and the organizational procedures that make data a competitive advantage are lacking why is that the case because it's difficult to change operations especially for traditional and established firms so how can we think about the transformational process of becoming data driven almost every organization collects stores or has access to data but
often data lives in silos and is under control of specific organizational units it is impossible to use it at scale let alone extract value from it the data is simply stored but not analyzed so how can you change that well start with a few selected projects that demonstrate the value of analytics based decision making without making big organizational cultural shifts then in the next stage you can start to implement bigger architectural changes to optimize how the organization handles data for example there could be a centralized unit that is devoted to data science and analytics finally
we can think of organization almost as ai factories where everything is centered around data firms that fully made the transformation to become data driven the value of data in terms of quantity and quality can come from putting different pieces of information together to create a database with enough features and examples that vary in the cross-section and over time but does this mean that data as a rare resource can only exist when the firm fully controls every aspect of it of course not in the extreme data can be almost like a commodity in a sense that
it's publicly available but it can still contribute to generating a competitive advantage in fact many firms use so-called application programming interfaces apis that give access to data from external firms for example to build applications on top of an existing architecture or platform as with any make or buy decision there is a trade-off between keeping data rare and non-immutable and between opening up for valuable inputs from outside the own organizational boundaries evidence shows that firms that use apis are more valuable on the stock market and invest less than own research and development but they are also
at greater risk of suffering damages from data breaches let me sum up we have clarified that data should not be seen as a resource in itself but always in the context of a larger value chain we have looked at a model of the organizational transformation to becoming data driven and went over facebook's strategic approach to data collection
