构建数据产品作为竞争优势

Contents1.0 Introduction2.O Project vs product approach to data3.OData products in action4.0 Lifecycle of a data product 5.0Data productmanagementa new function6.0The right skills and technology to createandmanagedataproducts7.0Summary Introductionsurvive and thrive in today's fast-paced business environment, That's whyData is a powerful new form of capital. It's essential for organizations towe believe that every CEO should be as strategic about acquiring.growing. refining, safeguarding and deploying their data as they are abouttraditional business capital. Today, data is generatecd everywhere. It comesfrom humans, machines, Internet of Things (loT) devices, edge systemsand morehere on earth and even in space. All that data scds up. In fact, astaggering 2.5 quintillion bytes of data are generatecl esch day.data-driven, The 2021 Chief Data Officer (CDO) Surwey by Arccenture andCEOs who see data as capital can grow and reinvent their lbusinesses to beMIT:, 78% of CDOs told us that their most critical responsibilities are drivingDusiness growth and value creationmaximizing the value of your data and treating it as an asset differentiatedSo, what does it take to become a data-driven enterprise? It means by its completeness, maximizing the potential value, lineage, quality, anddata and analytics, instead of tribal knowledge In short, it means thinkingease of use. It means basing critical business decisions on high-qualityof your data, and the analytics that rely on it, as a 'product"one that'sessential for driving business outcomes.This raises another question: What is a data product? Simply put, it's *aproduct that facilitates an end-goal through the use of data-, It's datapackaged with everything someone needs to understand that data anduse it to solve a new use case. The person can come from different teamor even fron outside the business. Just like consumer products, datawhat data products are and see sorne examples. We'll also explore howproducts are designed for specific purposes, Let's take a closer look atdata products differ from traditional approaches to using data. •••••••Projectvssproductapproachto dataToday, most companies apprcech enterprise datause cases. Data products can be:function has a specific problem that it wants to solvewith 8 project mindsot. In other worcs, a business• Datareusable datasets, streaming, feeds, and APlswith data. It then starts a project to acquire the data(Eg for design, manufacturing. finanoe andfrom the relevant sources, cleanse and prepare it.operations) that meet the needs of the wholethen analyze it for the specific use case. And eachenterprise, as well as each business function.time a new business problem or use case arises, it• Codefeature coce and transformation snippets(small blocks of reusable code) or data modelsanalyze data for its specific need.(representations of the logical structure of howdifferent data elements fit together).This project mindset may sound familiar. It's howmost data teams currently operate. But this approach• Analytics modelsreusable machine learninghas some significant drawbacks:models (eg for preclictive maintenance)• It may lead to duplicated work• Dashboard reportsreusable dashboards and• It's relatively slowother visualizations (eg a risk dashboard).• The outputs from each project typically can't beDate products typically offer far greater ROI andrepurposed to solve other use caseslower cost-per-use than clata projects. Why? Becauseevolve to support multiple outcomes over time andalthough the upfront costs msy be higher, theyData products provide companies a better way toaddress their data needs. First, the data teamproduct mindset keeps the focus on relizing theaccommodate new use cases that emerge. Thelooking at individual business units and across thedetermines the organization's data requirements,business use cases. They offer an innovative way toenterprise. Then, they create reusable data productsdecouple data from specific applications and usethat help multiple business functions solve differentcases to maxirmize its value. And they help breakdown data silos across the enterprise.Transform data into a competitive differentiator Data products also offer several key advantages forthe people who use them. These include:Speeddata product is pre-built. That means peopleTirne-to-insight is much quicker, because thedon't have to start a new project each timethey want some data.TrustUsers know that data products have gonethrough rigorous quality controlReal-timeinformationUnlilke static datasets, data products proviclereal-tirme data for decision-making.AccessibilityThe relevant clata is alreacly available, sopeople don't have to go and ask anotheteam for it.UsabilityA well-defined data product with a wll-defined interface is much easier to consumethan a raw dataset.Transform data into a competitive differentiator Dataproductsin actionPowering digital nativesYou may not realize it, but you're probably already familiar with some