This is a guest post by Darrel Raynor, President/CEO, Data Analysis & Results, Inc.
In today’s face-paced workplace, we are all asked to do more with less. Is that really the most efficient strategy? Here we define the resource allocation problem along with its myths. We then discuss parts of the challenge and how to address each part to relieve the stress of having too much to do!
What is Resource Allocation and Why is it a Problem?
Resource allocation is simply the assigning of people to operations work, project work, and special work. So many of our clients think they have a project resource allocation problem. They often ask for a tool or process to solve that problem. They usually want to evaluate software and purchase quickly. However, resource shortages are a symptom of resource over allocation, which is a symptom of either poor estimating or simply allowing too many projects to run concurrently. It can also be both! It is frequently a very political situation.
As always, we advise, “Set, test, and improve your process FIRST, then map your process, workflow, technical environment, and budget to a tool.” An ineffective process using a tool is still an ineffective process, it just costs more. We have seen lots of tool purchases that end up as shelfware, so don’t be in that group! Tools can be very helpful, once you figure out what your best practices should be.
Let’s review each of these common resource allocation issues so you can determine the best course of action, based on your situation.
Resource Allocation Problem
You have a resource allocation problem only if your people are working on the wrong things, not if they are spread too thin or lack the tools or dependencies to work on the right efforts. True resource allocation issues can be solved by grinding through the political process of prioritizing all efforts, posting the official prioritized list internally, and sticking to that vertical prioritized list. It will be obvious to everyone what they should work on next.
Executives, step up! Set your organization priorities at each level of management with accountability. Leaving the decision on what to work on to every professional in your organization is a recipe for anarchy or at least competing factions, and that is exactly what executives do not want. You may be amazed at the reduction in cross-functional pressure to perform when staff can point to a higher priority effort they are working on.
Resource shortages happen by poor estimating. Poor estimating occurs due to three main causes:
- Lack of risk-based and data-based contingency. We have a lot of experience with clients who simply do not want to analyze, document, and use contingency to avoid Resource Shortfalls. Teams usually have or can get data on both actuals and risks, it is just not encouraged for them to use their data! This is easily fixed by moving to actuals-based scheduling and changing just one word in the estimating game. Instead of saying, “How long will this take?” ask, “How long did this take last time?” Unless an effort is really ‘Top Ten’ priority, tasks will likely take just about the average of the last 3-4 similar tasks’ actual times. This can also be fixed by adding quite a bit of contingency up front, then reducing for future projects or tasks as you gather actual data. Timesheets and analysis will help.
- Asking people to estimate tasks where they do not have the skill or do not control dependencies. This is easily fixed by having people who understand the work and the skills of the people who will perform the tasks. This works best when staff is assigned to efforts full-time and do not have to juggle support, operations duties, ‘Tiger Team’ efforts, and other distractions. To improve estimates, add two factors L, M, H: Complexity and Skill. The easiest way to drive accurate estimates is to limit assigned staff to work on a particular effort full-time, at least until they finish or are blocked. Again, timesheets and analysis will help.
- People are not in control of their time. Here is a very frequent example. It is IT, but could be any ops/project multitasking situation. How can you expect people to give you a date certain when they can be redeployed at any time by a variety of circumstances?
Manager – “Please commit to getting your part of Project-A done by Friday latest!”
You – “Ok, you got it. I may have to put Project-B and other things off.”
Manager – “Ok, keep me posted.”[Later that same day…]
Manager – “I need a status on Project-B.”
You – “I had to put Project-B off for a week or so, due to upping priority of Project-A.”
Manager – “What!?! You know we have a commitment for Project-B!”
You – “Yes, but you said…”
Manager – “Well, they both have to be done by Friday and you knew that!”
And so on… Same thing when a customer support or executive request pops up.
Guess what? All of these situations can be helped by having data to use in estimates, and you get that data from timesheets and resource management.
Resource Overallocation – Too Many Concurrent Projects!
Resource overallocation occurs when assignments are given without really analyzing what is already on each staff member’s plate. When staff are given too many tasks to reasonably complete, they naturally can only work on some tasks, causing the other efforts to fall behind. The problem is compounded by well-meaning project managers, functional managers, and others who increase ‘pinging’ status meetings, and ‘transparent’ communications that can feel like public floggings. The Task-Switching Penalty goes into full swing, reducing work time by up to 20%! One great way to limit projects starting is not allowing them to start until they are fully resourced with time to work on the tasks publicly on everyone’s calendar. You know what’s coming next: Timesheets, resource management and analysis will help.
Timesheets and Analysis Will Help
Have you noticed a pattern? Our firm is Data Analysis & Results. That means we seek meaningful data to use in projects and operations planning, process improvement, and day to day management. Only when we know what is assigned to whom, how long similar tasks have taken in the past, and how often other things come up and how long they take, can we make meaningful changes to our rapidly spiraling estimate quagmire.
“Get thee some data!” we say to staff who are overworked to the point their work / life balance feels nonexistent. Armed with actuals, they have a much better chance of negotiating a reasonable task load and completion dates. Using contingency can reduce the required heroics of all-nighters, weekend work, and taking shortcuts, thereby increasing risk and a large queue of technical debt.
Moving to work negotiations based on priorities and tradeoffs, rather than one of just adding work without reason, will yield higher performance and employee satisfaction. Work can be reasonable if we use data to prove the time and effort tasks take. Let’s all be reasonable. There are fairly easy best practices out there, we just have to have the political will to use them!
Contact us if you want help with your culture of estimation, your tools, your processes, or just to bring some sanity to the situation!
Good luck and go home at a decent hour!
About the Author
Darrel Raynor serves as Fractional CxO to increase your organization’s efficiency, and decrease friction using clear communications internally, with customers, vendors, and partners. Darrel is a senior technology executive, speaker, consultant, trainer, and turnaround specialist with 30+ years of leadership experience streamlining operations, optimizing systems, leading projects, and coaching people. He has recovered projects and organizations and managed major new technology enterprise financial and other software development and integration efforts. Mr. Raynor and his associates serve as Fractional CxOs to drive revenue & alignment for owners, founders, Board, VC investors, or executives with a problem or in a period of high change. Tech, projects, SaaS, Merger Integrations, high growth, or Change Management. COO/CIO/CTO Start-ups to close-downs. DARaynor@DataAnalysis.com +1-512-850-4402 www.DataAnalysis.com