The concept of ‘self-service business intelligence (BI)’ started gaining momentum in the early 2000s. More than two decades later, a survey by Yellowfin has found that the majority of respondents (61%) say that less than 20% of their business users have access to self-service BI tools. Perhaps more concerning, 58% of those surveyed said less than 20% of people who do have access to self-service BI use the tool. In this blog, the first of a two-part series on the topic of self-service BI, I take a closer look into this interesting and challenging area in the wider data field and share my views.
Perhaps adding to the challenge is the vagueness of the definitions available surrounding self-service BI. A BARC paper describes it as ‘tasks that business users carry out themselves instead of passing them on to IT for fulfilment.’ Elsewhere, a paper on the history of self-service analytics and business intelligence defines self-service BI as the use of solutions that ‘have evolved beyond the limitations of traditional BI tools, offering a more intuitive and user-friendly experience.’
As with all things technology-related, the self-service BI segment has evolved considerably since the early 2000s. According to YellowFin, ‘self-service is no longer a term used to describe simply accessing reports. The definition of self-service has grown to capture data exploration, building reports, and doing your own analytics.’
The benefits
With the semantics out of the way, let me look at some of the benefits that self-service BI offers.
One of the main ones is the fact that it gives people autonomy from IT teams. Self-service analytics empowers end users to access and analyse data independently, reducing their reliance on IT teams. This shift not only frees up IT resources within the organisation but also enables faster decision-making and reduces the backlog of IT tickets.
Being able to perform self-service analytics also means that users can explore data from multiple angles. They can therefore ask their own questions and gain deeper insights as a result. The result is that data literacy across the organisation is enhanced, and a culture of data-driven decision-making is promoted.
Another advantage is that self-service analytics fosters a culture of agility and innovation. It enables users to quickly iterate, experiment, and explore new insights. This capability empowers organisations to swiftly respond to changing business requirements and stay ahead of their competitors.
Setting up such an environment requires an up-front investment (capex) to procure the tools, and establish the security, governance, and other capabilities necessary to run with this approach. Organisations can therefore reduce their opex by using a platform that serves all business functions and use cases thereby reducing the need for a large team of data analysts to support the business.
The challenges
Of course, self-service BI is not all smooth sailing. It is not as easy as simply getting a tool and letting a bunch of trained business users loose on it! As is evident by numerous studies, the adoption of self-service BI is very low. I can testify to that. In several of my recent client engagements, self-service BI has only been fully adopted by a small cohort of business users.
There are also data quality issues to consider. In a typical organisation, there are many data sources with multiple copies of the same data. In my experience, I find that the BI and IT teams mostly have a good grasp on which data to use. But how are business users supposed to know that? They are hardly ever trained on those systems. How do they know to include or exclude null values, depending on what they want to know? How are they supposed to learn about workarounds and exclusions, never mind which graph types to use for which business questions?
These are all things that do not come naturally. If a business user reports inaccurate information, and it is found out, it will take years to build the organisation’s confidence in the data again.
Self-service BI is easier to implement if an organisation has proper documentation and cataloguing of the data resources. Preferably, it must also have good semantic or business data models in place. I know there are few ‘model’ organisations out there that have their metadata, dictionaries, and business data models in place. Unfortunately, in many of the ones I have been working or consulting in, this aspect is still very much a work in progress.
A related aspect is governance and privacy. How are business users supposed to know who may have access to what information? In some industries like healthcare, for example, it is crucial to protect patients’ and clinicians’ privacy. In our BI team, we have strict data access policies and enforcement. These become much harder to enforce when business users are let loose on the data. Additionally, most companies battle with the siloes of information formed in various departments – with improperly controlled and governed self-service BI, this problem is just compounded!
Keep a look out for my next instalment next month as I will investigate the types of users that relate well to self-service BI, as well as steps to follow to make sure it is properly implemented.
