A lot goes into selecting a new BI tool. I’ve written a separate article covering some details around those considerations that can be accessed here.
Michiel van Staden, Data Analytics Lead, Absa
Michiel will be speaking at the Virtual Enterprise Data & Business Intelligence and Analytics Conference Europe 15-17 November 2021 on the subject, ‘Embedding a New BI Tool‘
These conferences are co-located with the Virtual Data Governance & Master Data Management Conference 15-17 November 2021. Click here to view the first article in series.
Once you’ve selected a tool or mix of tools to go with, development work can start. As per most projects these days, the general guidelines would be to not scope one massive waterfall project from the start.
Even if the tool theoretically seems to perfectly fit your requirements, you would still want to start with a proof of concept towards practically testing it out. Start small enough to quickly get a good gauge of the tool in action, whilst making sure to cover assessing all of your key capability needs.
Also don’t feel the need to stick with something if it just doesn’t deliver. Fail fast and then try differently.
Plug and play*
There should always be an asterisk next to terms like plug and play. I recently bought a desk, which took me a couple of days to figure out how to assemble.
With business intelligence tools, free versions can now sometimes be downloaded from the internet, but even there already, company firewalls and other security policies could complicate matters.
A fair bit of work will thus need to go into setting up the required infrastructure, fitting to your organisation.
A major piece of the puzzle will probably be around how to share the reports with relevant stakeholders. Do you need separate licenses for that and can it be done via cloud resources, or do you need to get on premise servers, specifically for this purpose.
You definitely do not want a situation where these reports are stored all over the place and manual workarounds need to be figured out for sharing. Ideally, if set up well, stakeholders will increasingly be able to easily go to one central place to access all relevant reports, whilst receiving relevant quick updates within their daily workflow.
How to automate is also key. In a perfect world all reports would just update automatically, with alerts for investigation, when something goes wrong.
Even with the latest and greatest tool, your business intelligence dashboards are not going to develop themselves.
The chances are that your BI, MI, data analysts or even data scientists might be used to just running reports on existing platforms, having inherited much and not necessarily developed reports or dashboards from scratch.
In case some of them have done development, it might be within very specific tool sets or coding languages. You do get BI tools that are easier to get the hang off, and this aspect is definitely key to consider and assess early on.
There is always the option to bring in consultants to give you a head start. Just make sure that what you get at the end of the project is production ready, i.e. a lot of work is not needed afterwards to make it work. Also push for skills transfer, so that you do not become dependent on contractors for every single small change going forward.
End User in the Loop
Sitting in a dark room for months trying to figure out by yourself what your stakeholders and end users need is never a good idea.
You want to involve a variety of end users of the BI from the start towards unpacking their exact needs and developing accordingly.
Often they won’t know exactly what they want or need. This is where a close working relationship with developers and some back and forth, test and learn is really useful.
In the end whatever is developed needs to serve the end users needs, whilst at the same time challenging them with relevant insight on new opportunities and better ways of working.
Having been named the 2020 Datacon Africa Data Analytics Leader of the Year by Corinium Global Intelligence, Michiel has 14 years in data analytics experience across fraud prevention, credit risk & operations to digital & marketing, dealing with very practical business realities and successfully guiding strategy through collaborative data storytelling. As a lifelong learner, he is currently embarking on his Masters in Big Data Science, whilst increasingly sharing his learnings via MentorCruise mentoring, conferences, podcasts & writings and also providing 1-on-1 coaching to help others progress on their journeys.
Copyright Michiel van Staden, Data Analytics Lead, Absa