An Introduction To AnalyticsOps
And Its Role In The Data Journey

Executive Summary

Today, companies tend to be data-driven. However, business executives consume dashboards, and rarely the raw data.

Data is ubiquitous, its quantity is growing exponentially. A few years ago, the need to extract, consolidate, enrich and test data led to the emergence of DataOps (“Data Operations”) to address data governance issues. DataOps has inherited the so-called “agile” methodologies that have proven their worth in the reliability of software development, well known as DevOps.

These days, while the costly task of data governance seems to be finalized or in the process of being finalized, trust in Analytics is still lacking.

During this time, Analytics has taken a critical place within companies, even for those producing financial reports or regulatory reporting. Strategic decisions are made by boards and management committees based on these dashboards.

With the rise of APIs and automation, KPIs from dashboards drive and influence the behavior of algorithms. What happens if your company publishes wrong KPIs, which can feed, for example, automated trading algorithms? Your company will be held liable, not to mention the negative impact on your brand reputation.

So what’s the problem? Let’s be completely honest: dashboards are simply not tested. It’s like testing each individual part of an airplane but never testing the plane once it’s assembled. A perilous approach, you will agree.

AnalyticsOps is a framework that prioritizes the development of automation and the consumption of analytics across a whole organization. It helps break siloes by bringing together data science, analytics teams, and the rest of the business through integrating, testing, deploying, monitoring, and adapting analytics in order to continually drive value to the business. 

This white paper will discuss the role and importance of AnalyticsOps for organizations to have confidence in the decisions they make.

Introduction To Data and Analytics In Today’s World

Data is everywhere. It’s not something we can physically touch but data is around us all the time. The use and importance of data are forever increasing, meaning more analytics thus more competition in the market. $25.5 billion were invested in Analytics in 2020¹ making it the top spend in IT. Moreover, according to a recent survey², knowledge workers spend more than five hours a day in analytics and 99% of them consider analytics essential when making business-critical decisions. However, this increased use of analytics for decision-makers still raises one big problem: analytics trust. Having one single version of the truth is Analytics leaders’ holy grail but many have failed to achieve it, with four out of five CEOs not trusting the data upon which they base their decisions³. Other barriers such as a fragmented market resulting in low adoption also add to the problem of analytics trust. 

At Wiiisdom, we believe that people should be able to make trusted decisions at all times when using analytics. For this to work, organizations need agile governance solutions such as AnalyticsOps solutions to tackle this problem of trust.

Analytics Quality: A Business Problem

Modern analytics allows users to use and consume analytics freely with their own data sources and in some use cases, without any input from IT. Allowing for freedom and employee empowerment are essential for innovation, but letting BI analysts create new analytics can create chaos with an organization. It increases the risk of running out of control and Shadow IT due to a lack of top-down control of heavyweight IT processes. 

Given the increase of modern analytics, finding the right balance between governance and self-service analytics is a key success factor. By its very nature, Analytics constantly evolve over time to keep up with changes in the market and user demands. Therefore, managers need to understand that anyone must be able to do self-service analytics in order to keep up with these growing market conditions. Creating trust brings higher usage and transforms companies into data-driven organizations. It also builds a data culture which is an important factor for successful self-service analytics; nonetheless, a challenge in itself because it requires collaboration from everyone in an organization.  

The Multi-BI Reality

In an ideal world, organizations would use one single enterprise BI and Analytics (BIA) platform but the reality is far from that. The analytics market is highly fragmented due to the highly competitive field of analytics vendors, each continually innovating and adding more capabilities for their customers. It is extremely difficult for one platform to cover all the needs of an organization, plus with mergers, acquisitions, and poor alignments between technology and business it’s near impossible for organizations to avoid using several BIA platforms. 

According to Forrester⁴, organizations use on average five BI and Analytics solutions, (this includes embedded analytics). In 2020, we conducted a survey and interviewed over 800 BIA professionals who recognized that multi-BI is the reality and is here to stay, with 3.8 BIA platforms on average deployed per organization. BIA professionals, organizations, and analytics vendors need to face up to this reality and understand there is this missing link for, as Forrester defines it, “BI Fabric Technologies” — technologies that integrate BI components from multiple BI solutions.

Understanding Data Governance and DataOps 

Before users are able to consume analytics on their dashboards and reports, there are various stages by which data moves from collection to use. This is called the Data Journey⁵. Every stage of the Data Journey impacts the following one, with the final impact being on the Data Analytics layer. During the Data Governance lifecycle stage, data engineers are responsible for creating data sets for analysts to integrate, which if done manually are very labor-intensive and error-prone. Thanks to DataOps technologies, data engineers are able to streamline the quality control, orchestration, and data operations aspects of their duties. This is achieved by automating tasks for agile development and testing/monitoring the data pipeline. DataOps focuses on a process-oriented methodology used by data teams to improve the quality of data analytics.

The Trend Of Automation

It is clear that an increase of data leads to more data analytics and thus higher usage, requiring more testing and development to ensure data trust. To handle this heavier workload and tackle the problem of trust, organizations need to automate this process. 

As a result of automation, data can be ready for people to consume much more quickly because the previous mundane IT tasks can be done automatically in a shorter amount of time. As Matt Turck explains in his recent 2021 MAD Landscape overview, people have been talking about automation for the past three years because it turns “analog companies that run on sending Excel spreadsheets around with a lot of manual processes, into companies that are intelligent and automated and run on data.⁶” This means organizations can easily adapt and react to the data presented. It is becoming the real superpower that enables data-driven decision-making.