Data democratisation is a modern concept that applies primarily to businesses and organisations.
The definition of data democratisation revolves around the principles of data access and visibility.
Data democratisation aims to enable wider data access and usability across a business or organisation. This allows teams and team members from various business touchpoints, teams and departments to tap into business data.
Traditionally speaking, data is siloed by specialists and accessible by those with specific skills and qualifications. This is what data democratisation seeks to eliminate.
In a data-democratised business:
- Brick-and-mortar store employees can access nationwide or international sales data, allowing them to dig down into what products are selling, to who and why
- Data from customer service departments can be used to tweak marketing campaigns, allowing marketing strategy to reflect customer feedback given via customer service channels
- Employees can view business reports on key strategies like sustainability, allowing them to feel party to some of the same data used by C-Level executives
Once upon a time, data would have likely taken the form of the humble spreadsheet (and in some cases, it still does, at least partially). Behemoth business spreadsheets would have been looked after and maintained by data practitioners. The data itself would have been accessible by few individuals and understood by even fewer. Data analysts would use that data to report KPIs and key statistics to those at higher business levels.
Data democratisation seeks to remove some of these barriers, enabling near-universal data access throughout a business.
The Challenges that Inspired Data Democratisation
Data democratisation did not emerge out of thin air. As they say, necessity is the mother of invention, so why was the concept of data democratisation invented?
As many will already understand, we live in an era of huge investment in data and its many associated concepts, technologies, strategies, methods, ideals and philosophies are now incredibly well-developed, wide and complex. Data is no longer exclusively useful to top-level executives and stakeholders – it’s a powerful business development and growth tool, not just a means of reporting KPIs.
Despite the hegemony of data in business today, Forrester estimated that some 87% of businesses and organisations have “low BI and analytics maturity”. This shows how data is usually siloed in ring-fenced, protected systems that are generally inaccessible across networks of employees and departments.
Some example issues of siloed or ring-fenced data include:
- Developing marketing strategies whilst being unable to access useful past data from another department.
- Being unable to access eCommerce data as an in-store sales manager and vice-versa. Poor access to omnichannel data.
- Performing analysis that is relevant to another department but being unable to grant them access.
Sure, you can contact the necessary relevant individuals and make an access request, or put forward a request to make the relevant data more freely accessible, but this takes time and such decisions are time-sensitive. The bigger the business gets, the harder this is to manage and the more time is wasted on bloated inter-departmental data sharing requests.
Plus, once you factor in the effect of, say, business mergers that see cross-relevant data siloed across different business departments from the same group, requests for data become even more complex.
Data is no longer the aristocratic, tiered monoculture that it once was, and big businesses have already made it simpler for employees to access organisation-wide data if and when they need to, according to McKinsey.
Case Study: Product Metrics That Matter – A Mixpanel Study
Mixpanel surveyed 160 product managers across India to discover what types of data they viewed as most useful, why, and what barriers prevented them from becoming more data-driven. Whilst data can be viewed as a panacea for business growth, some 7 out of 10 businesses cite that they’ve generally failed to create a data-driven organisation that emplaces data as a central concept within the business.
Mixpanel’s findings are continuous with this. Common themes as to why businesses found it difficult to adopt data-driven strategy were as follows:
- Data experts at my company are too busy to help me
- I can’t trust the data
- I don’t have access to the data I need
- I have access to data but lack the skills to find answers to questions
- The analytics tools my company provides aren’t designed for product teams
All of these points relate in some way to data democracy.
Theme 1 is a very common issue; say you’re given a task with a deadline where data can help, but you can’t get a response from IT or other data practitioners.
2 may happen when the data is in a poor state for use, either evidently unclean and rife with errors or hard to read.
3 is similar to 1 – when employees can readily access data useful to their task, there is no need to wait around for unnecessary permission.
4 relates to data literacy and training – employees need to be able to explore and use the data, access alone is not enough to achieve data democracy.
Finally, 5 relates here to product teams but is also relevant to other departments, the point being that one data analysis tool is not necessarily suitable for every purpose.
We can see that data democracy is not just about enhanced data visibility and accessibility throughout the business or organisation, it’s also about instilling a data-driven culture.
Employees should refer to the data before making important decisions. Reports and insights should be jargon-busted or visually formatted, allowing pretty much anyone to make good sense of data when completing a task for which that data is relevant and useful.
The Benefits of Data Democracy
Data democracy is fundamental to becoming a data-driven organisation.
By empowering individuals with data at multiple levels, a business can optimise its internal and external processes, cutting time wasted on data requests and allowing individuals to take the initiative. HBR also found that greater data literacy combined with self-service data tools helps employees, e.g. by enabling them to accomplish complex tasks quicker.
“Data democratisation means that everybody has access to data and there are no gatekeepers that create a bottleneck at the gateway to the data. The goal is to have anybody use data at any time to make decisions with no barriers to access or understanding,” – Bernard Marr.
We’ve already overviewed a couple of examples where this works, e.g. when a brick-and-mortar store manager is able to tap into regional, national or even international data to better plan their seasonal sales strategy, or when a marketing team wishes to use customer feedback siloed at customer services. Recruitment and HR teams may want to look at employee data to prevent employee churn and enhance onboarding.
There are many benefits to data democracy, though a lot of them may remain hard to forecast until a business takes steps to increase its employee’s access to data.
On the whole, data-democratic businesses are more efficient, lean, agile and boast better ROIs for their internal and external business strategies. McKinsey found that businesses that invest in creating a data-driven culture are more productive than their counterparts.
1: Data Discovery and Usability
Universal employee access to data enables that data to be used by more people. When someone has a query that can be answered relatively simply using data, e.g. “how many people bought a Christmas tree in this region on this day last year”, that data should be easily discovered and used. Data without access to those who need it is always limited – data democratisation allows data to disseminate into the hands of more individuals across a business or organisation.
This allows employees to solve their own data problems, not just for the business, but for themselves and their team. For example, say a type of product is running out of stock. An employee responsible for reordering that product would benefit from a breakdown of all the available suppliers.
This allows them to choose the best supplier, say if time is running short. In a data-democratised business, they can access supplier data that shows them what suppliers are capable of delivering how many products and in what space of time on average. This enables them to pick the best option, saving everyone time and stress.
2: Self-Service and Autonomy
As Bernard Marr highlighted, data-democratisation empowers individuals on a more individual level, hence its derivation from democracy itself. Say this same procurement employee has a hunch that X variation of an established product might sell. They can use self-service data tools to pull up potential suppliers, analyse pricing and run a small trial run on the product for comparison. They could then even compare results against the former product. If successful, their strategy could be replicated in other stores.
Whilst certain decisions or actions might still have to be sanctioned by managers, data-democratisation still allows employees to take the initiative and do their own research. This works for the good of the employee and of the business. Data is another tool that makes working life easier and more efficient – why should it be wholly secluded to certain individuals?
3: Creativity and Pragmatism
Data-democratised teams have the capability to be creative and pragmatic in solving problems. If they have a problem and theorise the solution can be located using data, democratised data tools allow them to source the evidence they need to find and prove that solution.
This helps foster a more localised and autonomous business model that is infused with self-responsibility. Instead of using tools and services sanctioned by another department, data-democratised teams can cut through bureaucratic inefficiencies and generate their own creative solutions to problems. Successful strategies and accomplishments can be broadcast throughout the company creating a culture of creative autonomy.
Concerns About Data Democratisation
Of course, we must be aware of the potential concerns and drawbacks of data democratisation. It resides on the cutting edge of data strategy and data governance and does not come without shortfalls.
Data democratisation in itself does not guarantee the responsible use of that data. Whilst businesses often collect vast quantities of various data that are traditionally siloed in various teams or departments, e.g. marketing, sales, customer relations, HR, etc, data democratisation relinquishes data from these silos.
This doesn’t necessarily mean that everyone has access to all data – that is the task of internal data governance and an inter/cross-department, top-to-bottom effort must be made to govern data in a data-democratised organisation.
Data governance becomes more complex in a data-democratised organisation and more stringent rules, regulations and policies will have to be put in place to ensure the security and proper use of data.
2: Data Literacy
Data democratisation works best when employees are data literate, or better, have expressed their interests in data democratisation already. Data democratisation is easier to justify if there are already evident issues with data usability and discoverability in the business, e.g. if employees know how, why, when and where to use data, but simply don’t have access to the tools and assets they require.
However, if employees are untrained in data or aren’t made aware of how, why, when and where to use it, then data democratisation requires more groundwork to come to fruition.
3: Time Inefficiency
Despite the potential efficiency of data democratisation, it might also cause teams to double up their efforts on the same things.
For example, you probably wouldn’t want a sales team to build up a customer profile in the same way as a marketing team – this is the job of marketing who can then share the data between departments. The doubling up of efforts is easy to avoid if employees, teams and departments check for pre-existing data products before deriving and creating their own.
The Key Components of Data Democracy
There are several technical and non-technical prerequisites to data democratisation.
- Integrations. Pretty much all modern data engineering for business involves integrating both internal and external data sources, data lakes and warehouses, analysis tools and platforms ranging from CDPs to CRMs, EHSs to OMSs. For example, integrating BI tools with SaaS cloud data platforms (like Snowflake) is vital to creating a cloud-accessible democratised self-service data tool for business intelligence.
- Defined input schemas and simplify metadata. Data mapping and terminology will differ between tools. Input schemas will have to be defined and delineated where possible to keep things simple for making cross-department queries. This is especially important where a multitude of tools or dashboards are being made available to teams that have never used them before – businesses need to establish a universally intelligble way to structure their data for querying.
- The catalog of datasets and tools itself. This is where employees, teams and departments can search for data tools and assets. Data warehouse providers, data analysis tools and business intelligence platforms often provide self-service data functionality that allows users to launch their own data queries. Catalogued self-service data products are how data democratisation typically works in a raw technical sense. More on this in the next section.
- Data literacy training and security. Teams and employees may or may not require training to use data tools and products, it really depends on their training and qualifications, etc. The better the business can break down and customise their catalogue of tools, the easier it’ll be for those with limited knowledge to access them. This also applies to working with data securely and in respect to privacy and regulation.
- Data visibility and awareness. Closely related to the above, a data-democratised business must spread awareness of this and its core principles. The aims of data access, transparency, efficiency and productivity should be spread throughout the business when embarking on a data democratisation strategy. Data democratisation is an opportunity to bring businesses closer together and this needs to be a core philosophy at the heart of data democratisation.
- Data governance. Data governance is intrinsic to data democracy. It concerns both the access to, integrity of and quality of the data. Access to different classes and forms of data will not usually be absolute from the top to the very bottom of the organisation, in the case of larger businesses at least. Establishing robust protocols for data governance is an essential pre-requisite for data democracy.
Breakdown of Possible Data Democratisation Tools and Services
Here, we’re going to give a rundown of a by no means exhaustive list of possible data democratisation tools and services.
Firstly, you’ll need to choose what tools to connect to your data warehouses, platforms and studios.
Mixpanel and Heap are both popular choices for product analytics whereas FullStory and Hotjar enable the collection and analysis of qualitative data. For A/B testing, VWO and AB Tasty are solid options. When it comes to onboarding, Userpilot and Userflow are excellent whereas Intercom is a superb customer service bots and chat service with plenty of data features.
It’s then about moving this data to somewhere where it can be transformed, categorised and optimised for data democratisation.
In terms of data warehousing (DWH) and business intelligence (BI), Snowflake and BigQuery are two of the market leaders, both of which provide seamless cloud integration with tons of other tools and services, data governance functionality and much more. Twitter was recently using BigQuery to democratise their general SQL analysis.
In terms of business intelligence (BI), Looker, Alation and Mode can work with a DWH for self-service analytics. Stitch or Fivetran can move data between systems via ETL and ELT data pipelines. Census and Hightouch are two other considerations for pumping data downstream. Visualisation tools including Tableau and Sisence help organise data in an easy-to-understand visual format for self-service users.
Once data has been centralised for organisation-wide governance, control and access, it’s time to train employees in data democratisation and slowly roll out functionality.
Data Democratisation In Action
Whilst many of these tools feature their own opportunities for self-service data analysis, the issue of governance enhances the appeal of a singular platform option like Alation. Alation is used by Cisco, who won the 2020 SuperNova Award for its data democratisation strategy.
Abra Le, Senior Manager, Change Management, at Cisco stated:
“At Cisco we strive to enable data democratization while simultaneously ensuring our data is properly governed. To succeed, we needed a platform that supported a people-first approach and provided business units with the ability to utilize data to drive their business forward. With Alation we consistently deliver high-quality governed data with context in a single platform, providing our business users with visibility into what data exists and where it resides, allowing them to generate more value from that data. We are honored to be named a winner in the Constellation SuperNova Awards for pursuing digital safety, governance and privacy” – Abra Le, Cisco.
See video below:
Of course, this is high-echelon data democratisation. Implementing smaller, more compact solutions is also certainly possible.
Cisco is probably the world leader in data democratisation right now, but numerous businesses are now working towards their own democratisation goals including Airbnb, which trained employees to become ‘citizen data scientists’. This cross-department data initiative allowed access to the brand’s analytics platforms and increased activity on those platforms by some 66%.
Pharmaceutical company Boehringer Ingelheim created a democratised framework of shared clinical trial metadata. This allowed researchers to tap into the same stream of universal data.
The Royal Bank of Scotland also created unified self-service data products in order to renovate their data culture and improve cross-department data interactivity; “Raising visibility from our digital marketing platform and data-driven strategies was vital to the shift,” Giles Richardson, the head of analytics, told InfoWorld.
Summary: Data Democratisation
Data democratisation revolves around providing a wider network of employees with the tools and training they require to discover and analyse data. Though simple in concept, data democratisation is fairly complex in practice, especially when there are many departments to consider.
Businesses looking to democratise should first consider who would benefit from greater access to data. What departments or individuals are frequently requesting greater access to data? Remember that data democratisation is mutually beneficial – both for the business and employees.
This is one of data democratisation’s primary advantages and shouldn’t be ignored.
From there, it’s about figuring out the best technical approach to data democratisation. In the vast, vast majority of circumstances, you’ll need well-oiled centralised data architecture to get started.
What is data democratisation?
In simple terms, data democratisation is extending the discoverability, access and use of data throughout an organisation or business. Instead of data remaining siloed in specialist departments, it’s made accessible by a diverse array of teams. This empowers employees across the business network with greater autonomy and control over data.
Why is data democratisation important?
Data is a valuable commodity and tool, but it’s not always intelligible or usable and can be quite esoteric. Data democratisation attempts to make data more usable for a wider array of non-specialist individuals, e.g. through the use of data visualisation platforms. This makes data easier to discover, access and use.
How do you democratize data?
Data democratization first relies on strong data foundations. The process of engineering pipelines between data tools, services and platforms are critical. The aim is to unite data under one easily governed data platform or data warehouse (DWH). From there, employees can access data insights by browsing data products from a self-service catalogue.