A data warehouse can take many forms – for example as a simple collection of SQL databases, or as a mixture of unstructured and structured data. Therefore there is no “click and run” provider, and each data warehouse needs to be tailored to the requirements. For example, just “dumping” all company data into one huge database will both not improve load times, and also not solve the problem that no one has an idea of what the data could be used for!

This is a mistake many companies are doing – just “collecting” data to say one is doing “Big Data” does not provide any benefits. Employees need to be enabled to easily work with data, and be provided with tools which solve problems they are not familiar with (autoscaling, storage, backups, …) for them, such that they can focus on their main expertise.

Many consultants will directly quote millions in order to scrap the whole existing landscape and start anew – this is not always the best approach (except if the consultant wants to cash in ;)). Rather the existing bottlenecks and possible improvements from a business perspective should be observed and a solution provided which solves these bottlenecks.

An example: If no one in the company is able to make sense of the data, it does not matter if you can query it in 200 milliseconds, or 20 minutes!

Therefore, our approach is to usually start small, have a look at the different business requirements, and even start with a simple co-existing copy of relevant data, which does not compromise existing workloads. Then, an iterative solution can be worked out in a close cooperation between the end-users and us.

It can easily be overwhelming, especially in huge established systems. Sometimes, it can be quite easy to get started, in setting up a parallel system which does not touch existing workflows.

Do you already know what your “quick win” would be? Or are you not sure yourself? How about a free 15 minute consultation with us – no strings attached?