Contributed article by Amir Qureshi, Regional Customer Success Director, EMEA at ThoughtSpot
It’s hard to find a large enterprise that isn’t undertaking a large-scale digital transformation initiative today. Whether it’s the fear of disruption, the urgency to adapt to digital native customers or employees, or the need to improve business performance – most digital transformation efforts include or are centred on becoming data-driven.
One might rightly argue that business intelligence (BI) and analytics software have been around for decades. Why should the current wave of transformation be so different than its predecessors? In my experience working with large telcos, financial institutions, retail and advisory firms, it feels like we are moving in to an era where in a combination of improved data-literacy and maturing BI & Analytics technologies enable genuine ‘data democracies’.
In a data democracy:
EMPLOYEES at ALL LEVELS are EMPOWERED TO ACT on RELIABLE DATA that they can EASILY ACCESS, ON-DEMAND to have a POSITIVE IMPACT on the BUSINESS.
This definition may sound utopian in the enterprise world. However, in the consumer world this is commonplace. Searching for factual data has entered our bloodstream thanks to tools like Google, LinkedIn, Facebook, Amazon, Netflix, et al in the palm of our hand. Citizens are constantly accessing reliable data instantly, and increasingly making better and bigger decisions with that data.
As individuals, we are evolving to become more data literate. Irrespective of our age, we’re all acquiring that fresh millennial mind-set of searching through all available data, and persevering until we find something that improves the status quo. Enterprises are just beginning to realise the potential of this evolution. Combine this evolution with sensible policy changes and up-to-date BI tooling, and there is a tonne of value in enterprise data-stores waiting to be harnessed.
Some enterprises will adapt and harness this value, others will perish to digitally superior challengers (think Blockbuster and Netflix). Let’s unpack some of the legacy barriers that are preventing enterprises from genuinely democratising their data:
Fear that access to data = anarchy
Managers in enterprises often fear that (data) transparency and empowerment will lead to anarchy and chaos. There is a perception that employees don’t understand the data and will make expensive mistakes. It is bewildering that these same managers, in their daily lives as consumers, use tools like Google every day to make significant purchasing, financial, and lifestyle decisions. At times, they too make mistakes, and go on to correct them too. So why do they behave differently in corporate environments?
Data access and speed
Data delayed is data denied. The value of data diminishes with time. If it takes time for end-users to access the right granularity of data required for decision-making, employees working at the coalface will not be able to solve problems as they arise. This ‘last mile’ of analytics adds real, tangible value to an organisation. When people don’t get data in time, they act either on their gut-feeling or on bad data, or do nothing. These behaviours can quickly get embedded in an enterprise taking it quickly down the road to mediocracy and eventually to an existential crisis.
Mistrust in data reliability
When data is unreliable, it leads to various problems that can compromise data democracy. People who fear making mistakes often err on the side of caution. For example, a planner might decide to hold excessive ‘safety stock’ in inventory as insurance.
Others might seize on occasional data unreliability to discredit legitimate insights that don’t align well with their perception or established beliefs. A Microsoft-sponsored survey by the Economist Intelligence Unit suggested 19 percent of senior managers won’t act on data if they disagree with its conclusion.
The most common reaction to low or inconsistent data reliability is indifference. People will simply ignore the system and act on their gut, or ‘rogue’ data sets gathered informally.
The above three barriers to data democracy often result in top data talent leaving the firm and ending up with a digital challenger. This can result in a vicious cycle and further reduce the ability of an enterprise to transform. How do enterprises that urgently need to transform overcome these barriers?
Attack the root cause of the fear of anarchy…
In my observation working with big enterprises the root cause of the fear of anarchy lies in corporate inertia. Middle and senior management do little to challenge the rigid processes and policies that limit employees from accessing the data they need to excel (and not just stagnate) in their roles. Managers must be brave and critically evaluate their investment, prioritisation and decision making processes. These process and policy problems must not be masked and presented as data protection or data security risks. Modern BI tools today are extremely feature rich when it comes to enforcing data protection and security. Data access can easily be restricted to column and/or field levels for individual users. Furthermore, these technologies are easy to use and do not require little or no training.
…and make the argument for less anarchy
To make an honest argument for less anarchy, traditional organisations should acknowledge the insidious forms of anarchy they already conceal. A common occurrence is when employees aren’t given access to data, they find their own ways to get it through informal channels. They store it in spreadsheets, on laptops and devices, and often systematically use this data out of date, and out of context. Even employees who go through proper channels and processes often form habits to make decisions based on data from a past period like the last week or last month. This hidden anarchy results in systemic failures that appear to come out of nowhere – so called ‘black swans’ or more commonly, business underperformance.
Genuine data democracies provide access to accurate, granular and timely data to all employees and anyone in the trenches who makes daily, hourly or even minute-by-minute decisions. I’m seeing first-hand through my work with large enterprises that easy, non-bureaucratic access to data actually creates more order and improves outcomes. When empowered employees are given reliable data, they act responsibly, evaluate decisions more carefully and make better decisions.
Accelerate the “last mile”
Modern Business intelligence and analytic tools with completely new approaches are focusing on the problem of last mile analytics. This is great news for companies that want to empower their people. There are two technologies that I see as game-changing in last mile analytics: search and AI.
New analytics tools that use a search-driven interface mean that people who want to find data answers can do so with the same ease and speed as using Amazon or Google. This means people can get quick answers to simple natural language questions in near real-time to make smart day-to-day decisions. It also introduces a sense of exploration and, dare I say, fun, to the data discovery process. This continuous feedback and interactivity reinforces people’s use of analytics and helps to cultivate a data-driven, data-literate culture.
But what happens in the familiar scenario where people don’t know what to ask in the first place? This is where the smart AI-driven systems can assist people by identifying trends and outliers in data sets and propose searches. These proposed searches can even be tailored to individual preferences and roles.
Build a trustworthy data environment
Transparent data governance and the reliability of the BI tool are critical to building a trustworthy data environment.
Educating users and stakeholders on how the data is sourced, how it is refreshed and how roles and access privileges are setup boost transparency and build trust. I have observed time and again that user confidence in the insights is much higher when they understand the data governance, management and maintenance structures.
Additionally, modern BI tools also make it much easier for end users to understand their data lineage. Simple data labels and visuals of data sources, joins, roles, user privileges go a long way in reinforcing trust with the data. These tools also generate insights (in seconds) from very large volumes of data – 100s of terabytes with multiple billion rows or data points. This gives the end-users confidence that they are not building their forecasts (future business) on a handful of data points.
Last but not the least, data democratisation is not just about giving people the ability to generate reports. It is about enabling the workforce with right skills and tools, and empowering them to make decisions that ensure the business thrives in the digital era.