Prioritization - AWS Prescriptive Guidance


A key to understanding and analyzing a portfolio is aligning your dataset to your organization’s business and technical drivers. Align these drivers to the data elements you collect and use them to rank each application that you’re planning to migrate. Common drivers include increased agility, cost reduction, and operational resilience. If agility is a primary driver, you might look at how many deployments an application has had or will have in a year. You can then use that data point during prioritization. For cost reduction, you might look at the expected annual savings of right-sizing the environment. For resiliency, you might want a data point that represents the expected revenue lost per hour of downtime.

Decide what criteria to use

In this phase, you work with stakeholders to define the business and technical drivers to prioritize the migrating workloads. Set the migration workload priorities according to the impact the workloads have on your business and technical drivers. Select pilot migration candidates and first movers. These might not be the top applications or migration groups based on portfolio prioritization, because there are special criteria to ensure that the migration and operations teams benefit from pilot migration waves. You will learn lessons during every cutover, so it is typical to start with lower risk workloads.

Select 2–10 data points to prioritize your workloads. These data points should represent significant differences in how soon you will gain value from having the workload migrated so you gain more value sooner. A good example would be selecting the business criticality (for example, mission critical or important) if you want to start migrating to gain experience. A poor example would be selecting the business unit (for example, marketing, finance, or facilities) when there isn’t a clear driver to migrate one unit before another.

Decide how to use the criteria

Once you choose meaningful data points, determine a scoring scheme for each value of each data point. Assign higher scores for critically lower applications and prioritize them to migrate first.

After scoring values for each data point, you can compare each data point with the other. This optional step is so that you don’t need to worry about ensuring you have every value of every data point aligned exactly with your prioritization as it can become overwhelming to keep aligned as you iterate on assigning scores to the values.

To compare data points, use a multiplier for each data point. For example, you can differentiate a Business Critical data point from a Business Unit data point by doing the following:

0.2x 0.4x 0.6x 0.8x 1x
Business Unit Business Critical

In this example, the Business Critical scores would stay the same (multiplied by 1) and the Business Unit scores would be 60% of their assigned score (multiplied by 0.6). This indicates that workload Business Critical scores are more important than Business Unit scores.

Once you have assigned a score to each workload, look at the distribution of scores across the entire portfolio. The scores themselves don’t matter. It is the difference between scores that matters. For example, you might find that the top score is 8,000 and the bottom score is 800.

We recommend that you plot the scores out as a histogram, so you can verify that you have a good distribution. The ideal distribution will look like a standard bell curve, with a few very high priority workloads and a few very low priority workloads. The majority of workloads will be somewhere in the middle.


The point of this exercise is to understand which workloads are most valuable to migrate first. So, make sure that you don’t have the same or similar scores for most workloads because that would mean that everything has the same priority. We recommend that you look at what is at the top and bottom of the list and see if you agree. If you don’t generally agree, you might want to revisit the criteria you used to score the workloads.