Technology moves at a fast pace—blink and you’ll find yourself hopelessly clueless about the current state of the art. To stay abreast of the trends that are reshaping the data analytics space, we’ve assembled our researchers and consultants to give you their best guess about what will happen in 2020. Here are nine trends for the coming year. (See the accompanying infographic for a summary view of the predictions.) 

Prediction #1: Data Exchanges Are the Next Big Thing

2020 will be the year of the Data Exchange and none-too-soon. One of the top data pain points for data analytics teams has been finding, evaluating, purchasing, and integrating external data, whether from commercial data syndicators (e.g., Acxiom, Dun & Bradstreet, Foursquare) or open data repositories. A data exchange simplifies this process by making it as easy for enterprises to acquire third party data as it is for a teenager to purchase music on Apple iTunes. The cloud has opened the door to universal data exchanges by collecting tens of thousands of customers on a single multi-tenant platform. Not surprisingly, cloud providers are leading the way with Snowflake and Amazon as industry first-movers: both have recently announced public data exchanges. (Listen to our podcast with Justin Langseth of Snowflake as he describes the promise and mechanics of data exchanges.) Before this data goldrush takes full steam, however, data providers and consumers will need to gain assurances about exchanges secure private and sensitive information. 

- By Wayne Eckerson, President, Eckerson Group

Prediction #2: 2020 is the Year of the Graph

Graph concepts and technologies have been available for several years but often relegated to niche use cases. Prepare to see graph adoption accelerate rapidly as the power of graphs is recognized for a variety of use cases. The graph concept is quite simple, with all information expressed as simple relationship structures composed of edges and nodes. A node is an identifiable thing, such as a customer or a product. An edge is a relationship between two nodes. The power of this simple but elegant model is in the wide variety of use cases to which it can be applied.  Customer 360 is an obvious use case. Customer data is everywhere and understanding the relationships of customers with events, transactions, and other business entities are central to CRM. Compliance is a pressing business need and one for which graphs can readily manage relationships among regulations, policies, processes, and data. Perhaps most compelling is the potential of graphs for data management. As volume, variety, and complexity of the data resource expands, the potential for AI/ML to expose data relationships (to other data, to users and use cases, and to regulations) and for graphs to manage those relationships can ease many of the pains of data management, simplify data architectures, become the foundation of data fabric, and reduce the frequency at which we lift, shift, and copy data. Graph technology is mature. Widespread adoption is the next step. 

- By Dave Wells, Senior Analyst, Data Management, Eckerson Group

Prediction #3: The U.S. Government Will Craft Federal Data Privacy Regulations 

As we get ready for the California privacy regulations (CCPA) to go live next month, we are beginning to hear the term COPRA resonating from Washington, D.C. COPRA is the new Consumer Online Privacy Rights Act introduced this month by U.S. Sen. Maria Cantwell, D-Wash. As it is currently drafted, COPRA will not preempt state privacy laws; however, it will provide a common minimum standard set of rights for all residents of the United States. The approval of this law will reinforce the need for organizations to integrate privacy principles into almost every aspect of their operations. Data catalogs, privacy professionals and data-literate staff will be in high demand to help build awareness of what personal data is collected, consumed and shared; so, those accountable for protecting the data can ensure that it is managed appropriately.

- Sean Hewitt, Senior Consultant, Data Governance, Eckerson Group 

Prediction #4: The Hottest Job in 2020? Ontologist 

In most organizations, there is a growing disconnect between machine learning algorithms and corporate knowledge. This disconnect is an example of a more general problem in artificial intelligence (AI). As articulated by DARPA, when machine learning algorithms are not guided by relevant domain knowledge, whether for banking, insurance or automated driving, they can generate false outputs, sometimes catastrophic. It’s critical to connect algorithms and data processing with rich, inference-ready knowledge (a.k.a. ontologies) that is independent of any particular representation in databases, file systems or other data management systems. Leading companies, such as Amazon, Google, Facebook, and Nike, are posting job openings for ontologists. There is a particular interest in people with skills in semantic technologies, such as OWL, SPARQL, and knowledge graphs.  Some far-sighted companies that understand the importance of corporate knowledge might even hire a Chief Semantics Officer.   

- Erik Thomsen, Senior Consultant, Artificial Intelligence, Eckerson Group 

Prediction #5: Data Warehousing Becomes a Strategic Service

Data modernization and monetization are entering a new phase.  Enterprises are building industry-specific, shared-tenant Data Warehousing services to enable their clients and partners to consume data within standard or customizable data warehouse environments. For example, a Fortune 500 consulting firm is rolling out a DWaaS platform on which clients in the healthcare industry can run real-time reports on member enrollment, claims processing and other operations. A major farm equipment manufacturer is preparing a similar offering for dealerships across the United States, as is a major financial services firm for its clients globally.  This DWaaS trend underscores the degree to which large corporations have digitized their businesses and are now confidently pursuing strategic data monetization initiatives.  With a lot of help from cloud and data warehouse/ETL automation technologies, more and more enterprises will convert data warehousing from traditionally bottlenecked back-office operations into predictable and repeatable revenue engines.

- Kevin Petrie, Contributing Analyst

Prediction #6 - Lean, Agile and Product Management Approaches Go Mainstream

Digital transformation is disrupting every industry across the globe and business agility is now a business imperative. Enterprises must learn how to adapt quickly to increasingly rapid changes in technology and economic conditions to avoid extinction. Companies that embrace Lean, Agile, and DataOps development approaches and execute their analytics portfolio using a Product Management mindset will succeed. These Lean-Agile enterprises will deliver high-quality, innovative, and market-leading solutions in a timely manner. They will enjoy happier, more motivated employees, a significant acceleration in time to market, a sizable reduction in data defects, and large productivity gains. 

- Mike Lampa, Senior Consultant, Data Management, Eckerson Group

Prediction #7 - Data Vault Reaches a Tipping Point

Data Vault modeling techniques are going to hit a tipping point in 2020 where a plurality of projects that involve building or re-factoring the "hub" layer of a 3-tier data warehouse architecture will employ this modeling technique. It's been taking hold slowly and surely for quite some time but 2019 was the first year that I started hearing from many clients and class attendees that they are moving in this direction. Data Vault has picked up major steam in the last few years and is poised to start eclipsing traditional 3NF data models. It will become clear in 2020 that Data Vault is the most significant improvement in data modeling since the introduction of dimensional design. For those of you who want to dive further into this topic, I recommend reading Dan Linstedt’s Building a Scalable Data Warehouse with Data Vault 2.0, which offers the best description I’ve read about how to build a modern data warehouse. It’s now one of four references that sit on my desk, joining Chris Adamson’s Star Schema: The Complete Reference, Ken Collier's Agile Analytics, and Michael Blaha's Patterns of Data Modeling.

- Aaron Fuller, Senior Consultant, Data Management, Eckerson Group

Prediction #8 - Streaming Databases Will Become Popular 

Streaming databases are becoming increasingly powerful. New SQL-like query-languages for streaming data, connectors to most applications and systems, and new scalability features make streaming databases more suitable for enterprise solutions. Streaming technology will not only be found in real-time systems, but it will increasingly replace traditional ETL and data processing tools. Moreover, event-driven concepts will become standard practice when building analytics landscapes.

- Julian Ereth, senior analyst, Eckerson Group

Prediction #9 - Time-Series Analytics Automates Operational Insights

Time is the universal dimension that weaves through every analysis and business performance review. But most enterprise dashboards only show a snapshot of business activity with some trending information. But that is changing. New time-series analytic software tracks metrics continuously and uses artificial intelligence to detect anomalies. Enterprises will first use the tools to monitor events from internet-based machines, devices, and applications. But soon they’ll use it to continuously track tens of thousands of run-of-the-mill business metrics (e.g., sales, quotes, returns, profits across every dimension). At first, the software will simply identify anomalies and generate alerts. But soon enough, it will identify the root causes of those anomalies, suggest remediation strategies, and predict future aberrations. The power of time-series analytics will dramatically change the role of data analysts: their jobs will evolve from analyzing data and performance to instrumenting alerts and models. 

- Wayne Eckerson, President, Eckerson Group