I’ve met with quite a few potential Business Intelligence shoppers off late and the most common debate I’ve come across is whether to go for an on premise solution or with a cloud based solution. It is not uncommon for most organisations to struggle with making the right choice here. But there are definitely a few obvious considerations to look at before taking the leap to either solution.
Invariably most decision makers tend to seek opinions from friends, former colleagues or even other similar companies and then try to follow in their footsteps. This gives them a sense of security…..albeit a false sense of security.
Companies must not only look at their immediate business pains but also consider what they hope to achieve long term. The BI solution they choose must provide the flexibility and longevity they need as they continue to grow and change. This facilitates the need to look beyond a brand name or the most popular platform and into strategic data driven BI initiatives. Choosing to go with a cloud based Business Intelligence solution just because that’s the direction the global trend is taking is probably not a wise move.
Data is the “OIL” of today’s world. Holding data about businesses and individuals are big assets and probably worth millions of dollars. Just consider the world of social media. Yes, they have given us a very connected world and something to be benefited from, but more importantly it is a platform that encourages individuals and businesses to share data. The social media service providers’ net worth is not from the services they offer, but from the data they hold.
Where is your Core Data?
Some organizations already use ERP in the cloud and possibly also have much of their CRM and other data sources on the cloud. For such organisations a cloud based Business Intelligence solution might be a natural transition. It could be a lot easier to add on data visualization and analytics functions to their existing cloud offerings. But if your organisation has got distributed data, then you could potentially expect a whole new bunch of complications.
This could be either one of or a different combination of all of the below:
- All of your applications such as ERP, CRM, Payroll and any other data are all on premise.
- Possibly you have your ERP and payroll on Premise, but CRM is cloud based.
- Possibly you have your FMIS on the cloud but all other silo databases on premise
- Possibly you have your FMIS, Payroll and CRM on the cloud but Inventory and Assets on premise
For a complete business analysis using a BI tool, you would need to bring in data from all your data sources, irrespective of where they are. This is not impossible with a cloud based BI solution, but it can be challenging and might require complex data integration solutions and therefore cost a bit more than you think.
How will data integration take place?
For most good Business Intelligence solutions, data from multiple sources need to be extracted, transformed and loaded into a central repository. We usually call this a Data Warehouse. Identifying the requirements of such a data warehouse which is also on the cloud somewhere, and integration of the data warehouse with your core data such as your ERP, CRM, Payroll, Assets register, and probably other silo data sources are potential things to consider. In most cases, cloud solutions providers will offer connectors to most databases, but these could be at additional costs.
A cloud based Business Intelligence will usually require integration of both on premise data (Finance, Sales, Purchase, Assets, Inventory) and cloud based data (CRM, Payroll) and therefore will require a much more robust and reliable data integration solution.
Data Security and Privacy
Most companies have data security and privacy policies. When you opt to go with a cloud based BI solution, you give up having to manage data security and privacy internally. These responsibilities now reside with the cloud solutions provider. For most businesses, this might not be a big concern. However, if data security and privacy is of a huge concern to you, then probably a cloud based Business Intelligence solution is not ideal. Consider the recent headlines on CNN where apparently classified data from the US govt with highly complex levels of security was hacked into. Ordinary business related data on the cloud with much less security tools in comparison might be far more vulnerable. Again, most businesses don’t probably need complex levels of security and therefore a cloud based BI solution might suit them just fine.
Public Cloud or Private Cloud
This is an important point to consider, especially if you are dealing with sensitive data. With a public cloud, you would simply access your BI solution via an internet browser with end users having little to no knowledge on where the data resides and what the security parameters are. However, with a private cloud, though you would be accessing your BI applications via an internet browser, you would have control over managing your own data. Much of this responsibility would lie with your IT team. The costs for each of these options are different. Most organisations with business analytics needs would prefer to keep their data private and choose to go with a Private Cloud.
Opex or Capex
Finally, it is important to consider costs on the long term. Business Intelligence on the Cloud will fall under Opex (operating expense) and might appear to be very appealing initially. But over time this can prove to be way more expensive depending on other issues surrounding the implementation; such as data integration, data warehousing, scope and expansion of the total BI project. On the other hand, Capex (capital expense) might appear to be a bit high upfront, but you could potentially depreciate it over time. With Capex, you would own the software license and usually pay a one-time fee for the software and services. Your recurring cost usually is a small percentage of the software cost annually for ongoing support, updates and upgrades.
Most organisations seek BI solutions to gain more insight from their data that can drive them to make strategic business decisions that propel them towards success. Hence the need to take extra care when evaluating BI on the cloud or on premise. Yes, cloud service providers will most likely have pre-configured service options and infrastructure that can help you get off the ground much quicker, but it is important you do evaluate the level of managed services required in comparison to doing some of it yourselves. I would personally opt for a private cloud solution if I had to.
What would you do?