The momentum around the Qliks and Tableaus of the world is taking the space of analytics by force and we are all keen and scurrying to get on that bandwagon - Running our pilots, testing out the performances of these tools, checking its effectiveness on Big Data sources and of course, moving into visualizations.
But this time, let us focus on a few other relevant things regarding this technology adoption. How big is the change from the reports-oriented world to the business outcome-oriented world? What price do I have to pay to get on this bandwagon? These questions will help us build a perspective as to where we want to reach at the end of all the hard work.
Let us start with some of the key pointers -
- Is this a journey from traditional reports to dashboards and visualizations?
The answer is NO. If you are looking at getting Qlik, Tableau and the likes (Analytics and BI platforms) just for visualization, then you are overlooking the strongest power that these tools have to offer. The journey is about making Business Intelligence (BI) self-serviced. The power of these tools is how strongly and how contextually we can create a data model that can enable the business to create insights and stories on the fly. You would be amazed at the potential and the value you can create with this approach.
- What price do I have to pay to get on-board?
The price you pay is your 'Past'(a change in business processes). If we are assuming all the previous reports and investments in databases, models and cubes will go waste, then we are not right. However big the transformation, operational reports will always exist and the trick should be to move them to a lower CapEx setup (maybe shared Clouds). And the databases can always be used as an intermediate presentation layer between your data lakes and self-service BI tools. This will help performance and of course, planning.
- Should I go for a subscription-based self-service BI setup?
There are two schools of thought. One that says, move your enterprise reporting to the Cloud and have the self-serviced on premise. The other says, have the enterprise system as is and move the self-serviced one to the Cloud. In essence, both have their pros and cons and asks for a conscious call. The latter is a good option for those who are not sure how they will scale with time.
- Do I really need to check the performance of in-memory BI tools?
QIX 2 and VizQL are quite powerful and optimized. They do a range of stuff like partitioning data, compressing it, indexing it and even optimizing the cache loads. Do not attempt to test the depth of the pond from the banks. Always identify a use case first, build the business perspective and then get only the data that you will use for this set to run your tests. Of course, you can work on the concurrency of users through LoadRunner or some other tool. This gives you the best approach to test the calibre of in-memory BI tools.
- Will business still have to depend on IT?
The answer is 'YES'. This adoption journey is a great opportunity for everyone to get on the same boat. The IT team creates models that enable business to do analysis of their data, adding perspectives to their immense experience in the domain. Consider a situation where IT has no understanding of the business. Obviously, the whole structure will break down. So the way ahead should be that, at the start, IT will be dependent on business and in due course of time, business will depend on business. That may be confusing, but basically what I mean is, everyone will be part of the business.
These are only a few points that we need to consider while we are on the path of adopting analytics and BI platforms. There is a lot more that we can look into and a lot more we can achieve.