Kathy Iberle (Hewlett-Packard) and Sue Bartlett (IIS/STEP Technology) have developed a model to determine the ratio of software testers to software developers. The following comes from the abstract of their paper “Estimating Tester to Developer Ratios (or Not)”.
Test managers often need to make an initial estimate of the number of people that will be required to test a particular product, before the information or the time to do a detailed task breakdown is available. One piece of data that is almost always available is the number of developers that are or will be working on the project in question. Common sense suggests that there is a relationship between the number of testers and the number of developers. This article presents a model that can be used in describing that relationship. It is a heuristic method for predicting a ratio of testers to developers on future projects. The method uses the model to predict differences from a baseline project. A reader with some projects behind her will be able to come up with a rule-of-thumb model to suit her most common situations, to understand when the model might not give accurate answers and what additional factors might need to be taken into consideration.
In the paper the authors present two case studies: (1) “MergoApp”, a e-commerce website where the tester-developer ratio was 1:4, and (2)“DataApp”, a database application to replace an Excel application, where the actual tester-developer ratio was 4:8. A copy of their model can be found at Kathy Iberle’s web site (http://www.kiberle.com/articles.htm). In addition, slides for the presentation can be found here: Estimate Slides.
Read Full Post »
This is the most comprehensive study that I have seen on IT Staffing. It was written by Lon D. Gowen, Ph.D., Lead Systems and Software Engineer at the MITRE Corporation. The paper is entitled “Predicting Staffing Sizes for Maintaining Computer-Networking Infrastructures”. It was published in 2000.
In this paper Gowan presents benchmark data for the “Number of Users Per FTE of CNI Support” for
- Systems Administration
- Help Desk
- Configuration Management
These numbers were observed in both a DoD and a Private sector enviroment. Acutal numbers are presented in the paper and are compared to numbers predicted using a COTS modeling tool.
Full paper available at: Predicting Staffing Levels
Read Full Post »
Posted in analyst, helpdesk, model, personel, programmer, spreadsheet, staffing, system admin, tech support, tagged spreadsheet, staffing on December 9, 2009 |
Leave a Comment »
TechRepublic has provided a spreadsheet model to help you determine the right number of people in your IT shop. In their approach, they focus on one category of staff at a time. In each category, a few key questions are used to focus the analysis:
- Is there a separation between application development and support?
- Do programmers work on multiple business applications?
- Are external clients supported, and if so, is there customized code for individual clients?
- Are major investments needed in software development of critical business applications?
- How big is the programming backlog, and what type of changes are being requested?
- Can the key business processes be accomplished better and more economically with a third-party solution?
- If you have external clients, can programmers be dedicated to and billed to specific clients?
2. Business application analysts and trainers
- Are new applications planned?
- Does the company support the installation of software for external clients?
- How knowledgeable are the departments and clients in the use of their business applications?
3. Help Desk specialists
- Are infrastructure calls separated from business application calls, or is the Help Desk support functions for both combined?
- How responsive do you need to be?
- Does the Help Desk have sound escalation procedures?
- What’s the level of client satisfaction for IT support?
- Is the response rate to solve user issues sufficient?
- Do you have a tracking system to monitor support calls, trends, and responsiveness?
- How many calls is the Help Desk handling now?
- What is the percentage of local users (as opposed to remote)?
- Does the company require 24/7 staffing of the Help Desk?
4. Network administrators
- Are major changes or enhancements planned/needed for the infrastructure?
- Is an experienced architect of the network in place?
- What has been the history of implementing infrastructure changes?
- Is there an infrastructure strategic plan?
- Is a change management process in place?
5. Desktop support specialists
- What are the company growth plans?
- Are major changes planned/needed in the desktop hardware/software?
- What is the percentage of remote users (other office buildings, cities, etc.)?
- Is the response rate to solve desktop issues sufficient?
6. Data Center operations staff
- Does the Data Center require 24/7 operation?
- What are the requirements for the Data Center?
- Is a “lights out” operation possible?
- Is the Data Center secure?
Using this model IT staff needs are based upon a number of factors, including the workload, anticipated needs, current capability of the staff, and maturity of the company. As much as possible, we try to quantify all the variables in each set of issues. Ultimately, it’s a judgment call based on the variable data, the level of support that you need to provide, and your experience in managing IT. If you can quantify the variables that affect levels of support, you’ll be much better equipped to determine your true needs.
A full description of this approach can be found at: http://articles.techrepublic.com.com/5100-10878_11-1061079.html?tag=e106
A copy of the IT Staffing Model in an Excel spreadsheet can be found here: http://downloads.techrepublic.com.com/5138-6321-730024.html
Read Full Post »
Coastal Technologies, developer of HELP!Desk software, has published guidelines for determining the number of technicians needed to staff a support center.
Establishing the proper support staff to customer ratio is essential for any organization. Each of the factors in the table below has an effect on the staff size. Assume a starting ratio 75 to 100 customers per analyst (75:1 to 100:1), then adjust the ratio for the following conditions:
||Support Staff Levels
|Experienced support staff
|Customer handling expertise
|Large number of products to support
|Multiple shifts and weekends
|Support staff possesses knowledge of the organization’s business
|Internal support only
|External support only
|Both internal and external support
|Service Level Agreements negotiated
|Multiple platforms to support (i.e. Web, PC, Mainframe, Mac)
|Automated tools in place
|Experienced support center management
|Support center has good reputation in company
|Center has bad reputation
|Multiple support center locations
|Quality Assurance or Quality Control responsibility
|Proactive support philosophy
|Support center has additional responsibilities not listed above
* = A staffing decrease may be required to meet your budget, but too small a staff can create bigger problems later on – dissatisfied customers, excessive stress, technician burnout and high turnover to name a few.
**= Initially, proactive support requires more staff per customer. The trend is reversed down the line as you should see a decrease in customer problems.
Read Full Post »
Posted in model, staffing, tagged model, staffing on December 3, 2008 |
Leave a Comment »
Staffing ratio rarely work effectively
Good advice from JamesRL. Staffing ratio is not a very effective model in today’s complex (not meaning necessarily big)opereations. When someone asks me what their staffing level should be I always say “depends”. There are always uniquenesses to an operation that makes it different and changes the mix. JamesRL gave a good indication of what some of those considerations are. The next best thing to a staffing ratio in the Gross Staffing Model. This model can be easly created in a spreadsheet and the process and calculation is as follows:
1. Gather the Necessary Information
Number of Incidents per year
Average Handle Time including After Call work time
Vacation, sick, holiday, breaks, training, projects, admin.
2. Determine Total Work Hours Required (TWHR) Per Year
[Number of Incidents per year] x [Average Incident Handle Time] = TWHR per year
3. Determine Work Hours Available (WHA)
[2080 Potential Work Hours (PWH)] – [Vacation, sick, holiday, breaks, training, projects, administrative time] = WHA
4. Determine Actual Work Hours Available (AWHA)
Determine the Utilization rate you will use for your model.
o A good industry average for an IT service desk is 75%but should not exceed 80% or the group will be faced with issues of agent burnout.
AWHA = Work Hours Available (WHA) X Utilization Rate (UR)
5. Determine Gross Staffing Level (GSL)
• [TWHR] / [AWHA] = GSL (number of staff required)
The problem here is that this model does not take service levels into account (arrival rate). To improve the accuracy you need to go to an Erlang-C based model. There are a number of those on the market, including ours (www.radarinteractive.com) however for smaller numbers I would go with the Gross Staffing Model(GSM). The GSM is flexable as well as it can be used for staffing for various tasks including: Email requests, desk side, etc.
These concepts are covered in the HDI certification training we deliver.
Read Full Post »