Overcoming the Barriers to Self-Service BI


Overcoming the barriers to self-service BIIn previous posts I covered how to make self-service BI attractive to the users and aspects to consider when implementing self-service BI. In this post, I give my opinions on overcoming the barriers typically encountered when implementing self-service BI.

As with everything else, there are always barriers and challenges faced with new processes and systems. In fact, while researching this topic, I came across a statistic that made me realise how many companies are still struggling with the concept. A new survey from Logi Analytics indicated that merely 22% of business users can access and use self-service BI tools, which implies that just under 80% of companies are still struggling with the barriers of implementing self-service BI. This is a scary statistic.

Having given a definition of self-service BI in the blog posts referred to above, I will not waffle on that any further, but get straight into the barriers.


Most people think that that the biggest challenge with self-service BI is around selecting the correct toolsets and implementing appropriate solutions. However, in my opinion, people can be one of the biggest barriers too! Many business users are resistant to take up the facilities offered by self-service BI – either they are too scared to implement it into their work functions, or they are just plain resistant to change. So many prefer to use the old ‘safer’ approach of requesting a report from “IT” (it should actually read from the BI competency), and they are quite happy to complain when they have to wait too long or when they actually get what they inaccurately requested. However, this can be detrimental to the entire organisation – so they have to be encouraged to take the risk – and once they actually do this, they will quickly start to realise the benefits that self-service BI can offer.
The key approaches to overcome resistance to change are communication (internal BI PR), very thorough training (on both the tools and the actual data they will be using) and a strong program of support and mentorship. I still remember the business users of a Telco client having to go through self-service BI tool training on a holiday destination dataset – it just had no relevance to them at all. Train them on THEIR data! And do not think when you roll out the self-service BI toolset, that the BI competency team is suddenly free to up-skill themselves on business analytics, visualization or to get the data warehouse better populated! No, they have a very important role, and a great responsibility, to mentor and support the business users through the adoption and transition process.

Lack of timely results

A fact-based approach to solving business problems is typically iterative. As such, people in many organisations think that getting to the solution through self-service BI is too time-consuming. However, what they forget is that although self-service BI may take a bit longer, especially in the beginning, the in-depth understanding the business users develop in using the tool, and applying it to their underlying data is invaluable. The end-result is also usually a lot closer to what they wanted, than what they would have gotten if they specified it badly for “IT” to do.

A useful approach is to teach them elementary prototyping / agile principles. Approaches such as to develop a prototype quickly to test functionality, fail quick and fail fast, refine the solution through quick successive iterations, and so on, will teach them useful tool-oriented problem-solving methods.


Many organisations already have standard reports that required hours and hours of work to get to a useful format, with accurate business-related content. So to many people it would seem that self-service BI would be even more demanding and would take even longer for anyone to get right.

However, if organisations undersood what is meant by and intended for self-service BI first, take note of what it should be used for, the reporting time won’t take as long, the results will be more accurate, and they would soon realise its value. To put that in more concrete terms, self-service reporting was never intended to produce the CEO’s monthly enterprise report, nor the corporate sales dashboard – those must still be developed and delivered by the BI competency. Self-service BI is more rightfully intended for the sales manager, or his data analyst, to quickly churn out a report highlighting the stores where certain products don’t sell – i.e. a much smaller, dedicated and focussed application.

Big Data

As we are all aware, big data is still making waves all over the industry, with many IT professionals associating analytics with big data. If organisations still do not understand big data and its value, then they may certainly feel that implementing analytics is not worth the effort. However, some data science and exploratory analytics can reveal insights that are hidden within the company’s big data, and never mind the big data, even in the organisation’s existing “normal” structured data, which may be beneficial.

However, I would strongly advise – that is not the domain of self-service BI. So a good starting point for organisations is to make sure it has come to grips with the realities of their own known structured data first, and only provide access to that portion of their data for self-service BI. Big data analytics falls outside the scope of self-service BI. Leave the big data for the data scientists.

Concluding comments

So while it takes patience, teamwork and strategy to overcome the barriers to self-service BI, in my opinion and experience the opportunities far outweigh the challenges. In addition, organisations must realise that self-service BI cannot be adopted overnight. It has a big impact on the business users, so the change must be managed through hands-on data-focussed training, solid support and mentoring, and efficient communication.




Leave a Reply