This blog aims to provide some insight into Managing the Cloud with a Staffing model based on Demand vs Capacity. This model is already
known in the Managed Services world which until now had relatively lesser number of Business Critical applications. As more and more ERP & business
critical processes are moved to the Cloud the content will become more relevant.
The document may appear complex to you, however the level of detailing to address clear action points in case of a Staffing crisis requires
me to provide the level of description. So kindly read the Summary in case you are not interested in the detail.
(please note I am working on some more figures and
illustration to make it simpler to understand)
Typically the Managed Services business always thought of Managing there infrastructure and the live customers in a “Factory” like model.
When you label an IT hardware & applications maintenance team as a Factory what is implied in context of this blog are as follows – High Quality through
clear repeatable processes & low Total Cost of Ownership (TCO), Scalability and readiness to meet unexpected demand in case of a surge in customer number, a clear portfolio of services offered – that ensures all tasks are known (planned tasks) and only very few unplanned tasks.
TCO is a factor of all the cost involved in managing the Operations.
Typically TCO = TCO + TCOp (TCOp being Total Cost of Operations).
One of the biggest cost involved in a IT Factory TCOp, especially when the Cloud hosts Enterprise wide applications is the People.
This is where a Demand vs Capacity Staffing model comes to play. TCOp ∞ Capacity (TCOp is directly proportional to Capacity).
Demand is typically the rise in the number of customers which is good news, but it will result in natural need for more capacity and
result in high Cost of Operations. As the demand goes up you will need more people capacity. However the Capacity is not uniform across skill levels
Here is where the most important point comes into play. For a factory every single tasks under each portfolio should be clearly known.
For instance for a given Service Portfolio Block –
Planned Tasks = Tasks (T1 + T2 + T3 + T4 + T5…….Tn) of Priority (P1, P2, P3, P4) but not necessarily of the same order might exists.
You need to evaluate each task and the effort required for them in terms of total PDs (people days) and skill level. If the skill level is
high then you can only execute those tasks with very senior or high cost resources. The TCOp will be high for such tasks.
So let us assume that P1 = T1 + T3 + T5 + T8 + T10 that requires 24 PDs of work.
Similarly you have a collection of tasks grouped under P2, P3 and P4 with respective PDs of work for each. It is highly possible that the
PDs for tasks under P3/P4 maybe significantly lesser or more compared to P1, P2. Whatsoever there, it is important to clearly know the PDs for all priority.
One more important thing to note is though a collection of tasks are grouped to a given Priority, it is still required to know the
Priority of each tasks. So you will have P1, P2, P3……Pn, where n = number of tasks. However the first x Priority tasks will be grouped to P1,
next y tasks = P2, z tasks = P3 and n-(x+y+z) as P4.
The reason for doing this will come clear later in the blog.
Now we could easily split the Skills of the Capacity based on the PDs for each of the Priority tasks and plan Resource loading for a given set of Customers.
Resource Loading = PD (P1) + PD (P2) + PD (P3) + PD (P4)
Note - When the Factory goes live we will get more clarity on the PDs, Priorities, the capacity and even the demand and need to make further adjustments.
So now our Demand vs Capacity will look as below.
Scenario 1 –
Demand remains constant. Consider that our demand remains constant and there is no increase in the customer number. After the factory is gone live now we will need to reduce the TCO.
The factor in TCO = TCO + TCOp and the TCOp proportional to Capacity, so how do we now reduce the Capacity?
For this we will need to make sure the tasks that are currently part of P1 needs to be pushed to P2, P2 tasks to P3 and P3 to P4. However the balance need to maintain High Quality.
Hence review the tasks and push tasks that will not affect the Quality to the next level. For this you will need to make Reusable components, reusable codes and in general – Knowledge Repository, Tools and Accelerators. By building more accelerators and the knowledge repository for your Planned tasks you will now start pushing PDs from P1 to P2, and subsequent priority levels.
This way you will eventually bring down your TCO without affecting Quality.
Scenario 2 –
Demand increases significantly
If demand increases significantly you will need to again consider Scenario 1 approach and first try to increase P2, P3, P4 capacity by consistently pushing P1, P2 to create Knowledge Assets. This is essential and should be the mandatory attempt to meet the Demand.
Next task is to start loading resources in P1 after you are clear that no more tasks can be moved to P2 and similarly so for P2.
To handle the Demand, know every single planned tasks and the skill required for them, prioritize every single tasks and then load resources. In case of demand remaining constant, continuously lower TCOp by pushing P1, P2, P3 tasks to lower levels respectively. Be very clear that this
should be done keeping High Quality intact. When demand increases, again first attempt should be to push P1, P2, P3 tasks to lower levels and when this is no more possible, start loading resources at P1.
This approach will maintain Low TCO and High Quality with a Demand vs Capacity Staffing Model.