Buffers and endpoints
Fusion Online uses buffers and endpoints to support reliable project delivery. These concepts—simple at first glance, yet rich in their implications—help model and manage uncertainty while keeping teams focused on what matters most.
This article provides a thorough discussion of both endpoints and buffers, including how they’re created, what they represent, and how they differ from traditional approaches.
Endpoints: A simple but powerful concept
An endpoint represents a delivery commitment in your project. It could be the completion of the project itself or an intermediate milestone that’s important to stakeholders. In most projects, there is one primary endpoint, but there may also be a few others that represent critical interim deliverables.
Endpoints form the targets around which your project is scheduled. In critical chain scheduling, buffers are created to protect these endpoints against variation and risk. That protection is what gives the project its flexibility and resiliency.
Buffers: Managing uncertainty without micromanaging tasks
What is a project buffer?
A project buffer is a block of time inserted at the end of the chain of tasks leading to an endpoint. It is not part of the work itself—it’s a time allowance, designed to absorb the variation and risk that are naturally part of executing a project.
In the ProChain System, instead of artificially padding each task, we remove safety time from task durations and concentrate it into buffers, where it is visible, measurable, and manageable.
You can think of the buffer as creating a range of delivery dates:
The project is unlikely (say, 5% probability) to finish before using any buffer.
But there’s a high probability (say, 95%) that it will finish before the buffer is fully consumed.
That’s a more honest and useful forecast than a single due date attached to a long, fragile chain of task estimates.
What makes buffers in Fusion Online unique?
Fusion Online takes a more modern and integrated approach to buffering than traditional critical chain tools.
You may schedule multiple interconnected endpoints in a single project. Each can have its own critical chain and project buffer.
Chains of non-critical tasks feeding into the critical chain are scheduled early in order to protect the critical chain.
For those integration points where there is insufficient time to protect the critical chain tasks, Fusion Online calculates merge bias or integration risk. This is an estimate of how much to increase the project buffer to accommodate the variation contributed by each integration point.
When are buffers created in Fusion Online?
When you first schedule a project, Fusion calculates and inserts a project buffer for each endpoint. This marks the project as a critical chain schedule and enables status tracking tools like the Fever Chart.
Why might you reschedule?
You might choose to reschedule a project if:
The network has changed significantly
The buffers are no longer credible (e.g., a buffer is fully consumed, or not consumed at all)
A newly added endpoint needs to be buffered
You're implementing a recovery or acceleration strategy
You can trigger a reschedule from the Project Details view.
Can you re-buffer without rescheduling everything?
Instead of rescheduling the entire project, you can selectively re-buffer an individual endpoint:
Go to an endpoint view or the Endpoints tab of Project Details
Click the action menu next to an endpoint
Choose Rebuffer on Next Update
This will cause Fusion Online to recalculate the buffer size for that endpoint at the next schedule update, using the current buffer settings.
You can also delete or resize existing buffers from these views.
Calculating buffer sizes
Buffer sizing is central to the ProChain System. In Fusion Online, buffer size is based on:
Task variability along chains feeding the endpoint
Integration risk from merging paths
Project-specific buffer options you configure
You can control buffer sizing from the Project Options, found under the action menu in Project Details.
Buffer sizing options
There are three buffer sizing options:
Buffer Size (fixed duration)
A constant value added to each buffer
Default: 0 days
Buffer Size (%)
A percentage applied to the cumulative variability along chains feeding the buffer
Default: 50%
Remove Integration Risk (checkbox)
If enabled, subtracts calculated integration risk from the buffer
Rarely recommended
Buffers will not be created if both the percentage and fixed values are zero.
Variability and integration risk
Each buffer is made up of two main components:
1. Variability component
This reflects uncertainty in task durations.
For each task, variability = Low-Risk Duration – Focus Duration.
Fusion finds the chain of tasks with the highest cumulative variability.
The project buffer size is determined by taking the total of this variability value, multiplying it by the buffer percentage, adding in any fixed buffer component, and adding the integration risk component.
With various kinds of task links, it is possible that tasks can overlap one another. For example, a start-to-start link between tasks A and B with a lag of 2 days means that B must start at least two days after A. When there is overlapping, only the fraction of the predecessor’s variability that is NOT overlapped will be considered to contribute to the variability in the successor.
2. Integration risk component
This accounts for multiple chains feeding into a common successor. When several tasks are integrated, the probability of finishing all of them by a particular date goes down.
Fusion uses a proprietary algorithm that considers:
Number of feeding chains
Variability across each chain
Critical and non-critical resource dependencies
Task link types (e.g. start-to-start)
The more tasks that converge or share resources, the greater the integration risk.
Integration risk calculations are based on the project buffer percentage, so changing it from 50% to 25% in a simple network could cut integration risk in half. It might be cut by more if there were slack in the network.
If there's not enough slack time in feeding chains to absorb the integration risk, the shortfall is added to the project buffer.
What you can do with buffers
Once buffers are in place, they become central to project tracking and control.
Track progress and risk
The Fever Chart shows buffer consumption over time
The Impact Chain shows the critical chain feeding the endpoint
Tasks near the chain (shadow chains) can be monitored using the sensitivity filter
Update or resize buffers
You can recreate a buffer using Rebuffer on Next Update
You can resize a buffer manually by setting it to a specific size (e.g., to increase it based on new risks discovered)
You can also delete a buffer if needed
Buffers are visible and editable from:
Endpoints tab of Project Details
My Endpoints or custom Endpoint views
Project views
Viewing buffer status for multiple projects
This project view displays six active projects, each with its priority endpoint shown alongside the buffer status color and trend indicator. In the Gantt chart, the critical chain appears in pink, and the buffer is shown as a white box with diagonal cross-hatching at the end of the chain. The amount of buffer consumed is visually represented by how far the critical chain extends into the buffer area—making it easy to assess project risk at a glance.
Viewing the buffer status and projected landing zone
Fusion Online offers two types of fever charts: the traditional fever chart, which plots buffer consumption vs. percent complete, and the time chart, which we generally recommend for ongoing project analysis.
The time chart plots status over calendar time and gives a forward-looking view of project health. In this screenshot, the cone of uncertainty represents likely paths of the chart in the future, based on the current status and schedule logic; including a likely range of completion dates. The landing zone at the top shows where the endpoint is projected to finish, while the buffer area is demarcated by two vertical black lines. This visual makes it easy to interpret schedule risk and time remaining in the buffer, helping project managers focus on corrective action if needed.
When 50% isn’t the right answer
The default buffer percentage (50%) assumes that safety time has been removed from task durations and consolidated in the buffer. But there are exceptions:
Low-risk environments (e.g., well-defined construction projects with little multitasking): 30% may be sufficient
High-variability domains (e.g., R&D, new product development): 75–100% may be more appropriate
Always consider:
How much uncertainty is in your tasks?
Are task durations already conservative?
How costly is a late delivery?
The buffer is not just a safety net—it’s a signal. Right-sized buffers improve focus, reduce noise, and help you make better decisions.
Summary
Buffers and endpoints are essential elements of the ProChain System, and central to how Fusion Online models and manages projects. They provide structure, clarity, and a path to reliable execution—without the stress of artificial deadlines or invisible risk.
By defining clear endpoints and protecting them with realistic, data-driven buffers, Fusion enables project teams to track performance, absorb variation, and stay focused on delivery.
See also
Last updated
Was this helpful?