My firm ran some numbers for the organization and took a look at the profile of the data to evaluate the cost for solving the analytics problem.  Options considered included staying on the existing platform as well as lifting their existing data model and shifting it to an Oracle Data Warehouse Appliance (ODA) or AWS Redshift, otherwise known as a "Lift & Shift".  

One of the more surprising insights gleaned from our findings was with the Redshift Lift & Shift option.  The physical data model that would be used in Redshift could stay exactly the same as their existing data warehouse in terms of how it's interfaced with by existing consumer BI apps.  ETL into Redshift would need some, nominal, syntactical changes.  However, if the organization tried to stay on their existing Oracle technology to solve the analytics problem all consumer apps would need to change to experience the performance benefit.  Furthermore the database size and batch processes would increase another 50% on their existing Oracle  technology but be reduced by about 75% on Redshift.
True, Redshift isn't the only technology that could solve their problem this way.  The Oracle ODA option also had MPP and options to use compression.  But to stay with Oracle to solve this problem would cost about 4x more to do the lift and shift to ODA than a comparable one to AWS Redshift according to the increases licensing spend with Oracle over AWS Redshift and the inability to go columnar.  The cost of staying on the organization's traditional Oracle platform with SAN was looking like a 400% larger consulting/project spend to revamp and tune the existing solutions data model and ETL over keeping it the same.

No-brainer right?  That's when the dilemma kicks in.  When an organization is tasked with "Keeping the Lights On" as their first priority, the risk of making a change is not taken lightly.  This is compounded with the upfront sunk costs of infrastructure, licensing and resources to perform a Lift & Shift before any possible benefit is even experienced.   AWS certainly helps minimize this risk by taking away the long-term commitment to hardware and licensing of their products and even offering free trials.  However, to migrate 2TB from one server, data center and platform to another takes people and time.  That's a risk we find many of our customers unable to take.