Analysis Process Evaluation

There may be several reasons for a need for an evaluation of the analysis process within the organisation. One of the most typical is the moment where the need for different analysis coming from business is ever growing and at the same time the team of analysts are over worked and fully booked. Many times in these situations the analyst team is so busy that they can not stop even for a moment to re-plan the processes. And, another issue is that the ideas and fixes tend to be bound with the current way of doing things. Varjus Consulting services may be used to give an outsiders expert view on the process in these occasions. 

Quite a typical scenario is where the use of analytics in the company’s business processes has grown gradually over the time. The company has hired their first analyst and given him the freedom to build the needed analysis as he sees appropriate. And usually it is “quick and dirty”- approach “just to start with”. But, if the analysis has been helpful for the business the need and demand for additional analysis is starting to grow quite fast in the organisation. If the supporting IT- environment, roles and responsibilities are not planned well (or at all!), the company runs quite fast into a situation where the number of different analyses is limited by the throughput of that single analyst. And the typical fix is just to hire the next analyst and let him/her also build the needed analysis as s/he sees appropriate. 

The issues raised from this quite typical development scenario is that the analyses become dependent on the individual analysts. If the one analyst leaves the company the next analyst will need to redo the same analysis from scratch. There is no organizational learning. 

Also, if the supporting IT- environment is not designed appropriately there is no proper scaling for the analyses. The number of analysts creates the bottleneck for the number of business analysis that can be supported. Quite typical issues in the companies is that the analysts tend to take care of the whole process. They extract the needed data, they prepare the data for the analysis, they perform the analysis, they implement the model in production, they follow-up the model performance and once the performance is not good enough anymore, they re-do these tasks all over again for this one analysis. The fact that many analysts are skilled to do all these steps does not mean that it is the most efficient process.