1.    Introduction

Last month, I raised the following challenge:

How would you recommend addressing the following issue that is common in internal projects in enterprises that are not project-based:

 The objective is: “to use internal technical personnel as resources on internal improvement projects, without jeopardizing day-to-day operations”.

 The situation is as follows: projects need predictable availability of human resources; operations needs to be able to address unforeseeable situations immediately they occur; there is not enough project work to justify dedicating technical project resources.

 The conflict is as follows: projects need to be able to rely on advance resource planning with the required resources available for a given period at a predefined time; operations needs to have priority without notice for the same resources.

How would you analyse this in more detail so as to propose realistic solutions to allow projects to deliver reliably while relying to a certain extent on resources from Operations?

In Part 1 of the description of the solution, I proposed the staffing model shown in Figure 1, and proposed using the Erlang C model to evaluate the optimal number of resources to cater for Unplanned Remedial Actions (URA), for a given environment and defined service level. What remains to be explained is how to convert this theoretical approach into a working, operational reality.

Figure 1: The three categories of operations group resourcing needs

2.    Applying the New Model

We made considerable progress in part 1 in answering the first three of the four questions in section 2.1 of last month’s posting (see Figure 2). The answer to question 4 will depend on the financial model and the relative value of the actions in contention for URA resources. We now need to define how these resources will be applied to service the various calls on their time.

 Figure 2: The questions to be answered for using operations resources in projects

2.1    Servicing URA requests

The model assumes that you can accept to have asynchronous service requests (i.e. the URA requests) waiting in the queue on average for the specified time. However, it is not always possible to allow this since some requests may be too urgent to wait. In this case, a formal pre-emption scheme needs to be agreed in advance – e.g.:

1)     Take highest priority URA request.

2)     Identify the operations (SOA) activities to be delayed.

3)     Pre-empt the corresponding, previously-committed project resources

4)     If this is still not enough (this situation will be rare if the Erlang model has been correctly applied and regularly updated), then consider pre-empting KBR resources.

2.2    Staffing for Operational Staff in Projects

A financial model needs to be established which provides a viable approach for supporting project work with resources within the Operations Group. This should be based on a forecast of the number of forthcoming projects, their value to the company, and the involvement required from within the operations group. In addition, of course, project portfolio management must take into account the constraints imposed on SOA resources by this model and be responsible for forecasting and justifying the SOA staffing numbers.

2.3    Planning SOA resources

In order to make this staffing model work, the total load of SOA resources promised in any category to projects should not reduce the number of Operations’ resources available for URA activities to a level such that the average waiting time would exceed a pre-specified limit, except in cases where there is a joint agreement between the Project Sponsor and the Operations manager to accept this risk – with a given contingency plan.

3.    Implementing this Approach

3.1    Advantages of this approach

The advantages of this approach, of including the skilled project resources within the Operations Group include the following:

• exchange of experience within the team
• availability of additional resources for Operations when not required by projects
  • … and vice versa!
• a large set of operational resources to call on in real emergencies
• a larger number of resources will smooth out statistical exceptions
• more effective handover from projects into Operations.

3.2    Planning the implementation

This approach needs some reliable, quantitative data per resource skill category:

1)     the number of staff required to handle the KBR activities

2)     the number on SOA activities per week

a)     the average duration of each such activity

i)      or, better, a 3-point estimate

▪       shortest, longest, most likely values

3)     the number on URA calls per week

a)     the average duration of each such call

i)      or, better, a 3-point estimate

▪       shortest, longest, most likely values

It would also be interesting to see a cumulative histogram based on weekly figures of these three quantities plus a count of the number of times activities are delayed for each type of activity, and by how much. This second set of numbers will help to validate the validity of the model.

Once these numbers are known and the model validated, it can be extended to take into account forecast and committed project resources in each category so as to evaluate the number of additional resources required in each skill category to account for project needs without negatively affecting the operational service levels.

4.    Organisational Change Programme

These ideas should be discussed and, if agreed, integrated into an organizational change initiative in order to ensure a consistent and smooth company-wide adoption of the concepts and capabilities – a project that needs by definition to involve staff from operations!