Chris is the passion and brains behind Carv. He heads up product and oversees the egineering team.
Why it matters
Do you have the right mix of resources for tomorrow’s projects? How about next week’s or even next month’s? As I was building Knectar it always felt the like the answer was “I don’t know,” instead of “of course”. This has always been the biggest gripe running my own business and as it turns out, is a problem shared by the vast majority of service companies like mine. For companies executing projects on behalf of clients (e.g. web development, marketing agencies, consultants and CPAs), this uncertainty is a fundamental business problem.
Unlike larger corporations, agencies and professional services firms cannot solve this problem by adding additional layers of management to distribute work to individual contributors. We have a need to minimize overhead, and every FTE has a huge impact on our operating margins. Nonetheless, the problem is so acute that a new role “Resource Manager” (or “Traffic Coordinator”) has arisen to solve it. Our research suggests this becomes a full-time job at surprisingly low head counts. I’ve seen agencies with as few as fifteen people requiring a full-time resource manager in addition to their line up of project manager.
This is occurring because our current tools force us into Gantt chart level specifics to model capacity needs for future projects.
Think about that statement and the information you have to estimate future resource needs: budget, hours per role, and timelines. These are the most important aspects of modeling resource needs and can be estimated with reasonable accuracy on the fly. As a manager, you know roughly how much design work vs development work will be needed. The shortfall arises when attempting to aggregate estimates for all projects, prove it, and communicate it to the team.
We solve this problem with the best tools available: spreadsheets, Gantt charts, SAAS scheduling tools, etc. Unlike our estimates which are top-down, these project management tools require a bottom-up analysis of a project to provide value. This is fantastic for tactical project planning and execution. Their ability to break projects into individual tasks, dependencies, start dates and work-hours is an asset for scheduling and work allocation, but is a liability for long-term resource planning and project modeling. They force us into the weeds to define specific project variables too early in the process, and the additional time does not result in more accurate strategic resource planning.
Plan future projects, but plan quickly
Our experience from growing Knectar and working with early Carv users has confirmed that resources, budget, and timelines are the three levers that can be used when planning for future projects and negotiating contracts. In our early days at Knectar we estimated values for these, which resulted in inaccurate measurements of the tradeoffs between project and company level metrics. For example, we knew we could increase the budget and shorten a project’s timeline, but did not have enough data to determine the impact on resourcing and margins. This resulted in resource pinches, erratic margins and sleepless nights.
For the past six years we’ve been working on a solution to fill the data gaps so we could connect our project pipeline to implementation teams and solve our resource management problem (today we call it Carv). I am now at a point in the process where I want to share my approach so other services companies can solve their resource management problems.
My approach to strategic resource planning:
Define the budget, start date, and the number of hours needed from each type of resource/skill to execute a project. For example: 50 hours of development, 65 hours of design, 25 hours of project management.
My estimates for Williston:
Map the estimates on a timeline based on weekly workloads and account for how/when the roles will actually be contributing to the project. For example: I account for ramp up/down time and when roles will start contributing to the project. A project manager (blue curve) will typically start working on a project before the developers (purple)
Williston in the project modeler
Repeat steps 1 and 2 every time new projects are added to my pipeline. I add projects when I can make ballpark estimates for workload per role. The modeling takes about 5 minutes.
Map all the projects on the same timeline and look for resourcing gaps. With the tool I built, this now happens automatically! Before I did this manually in Excel and it did not look this good.
When a project is confirmed and nears its start date (four weeks out for me), provide the project model to my project manager as a starting point for their tactical Gantt charts. I have found using this as a starting point has simplified the project manager’s discovery process and increased accuracy of the project schedule.
I’ve been using this approach and tool at Knectar for years and it’s served us well. We’ve honed the process over the years and my friends/colleagues finally convinced me to open it up to the public. The screenshots in this post are actually from our live Beta! The success I have had is already translating into a more predictable resourcing model for a handful of early beta customers. If you want to give it a try, get in touch!
Get in touch:
Chris Amato @ firstname.lastname@example.org