S-curve in project management: what it is and how to read one

What an S-curve is

An S-curve plots a cumulative quantity on the vertical axis against time on the horizontal axis. The quantity can be money (the most common), labor hours, or percent complete. Because it is cumulative, the line only ever goes up or stays flat.

The S shape comes from how most projects use resources. At the start, only a few activities run: mobilization, design, procurement. In the middle, many activities run in parallel and spending per month peaks. Near the end, only finishing, testing and handover remain. Monthly spend therefore looks like a hill, and the running total of a hill looks like an S: flat, then steep, then flat again.

Not every project gives a neat S. A software subscription bought up front, or a project with one large equipment purchase, creates a step. A team of fixed size working at a steady pace gives a nearly straight line. The shape is a consequence of the plan, not a rule the plan must follow.

The three curves: PV, EV and AC

A single planned S-curve is a forecast. It becomes a control tool when you add two more lines, the core of earned value management:

• Planned value (PV): the cumulative budget for the work scheduled to be done by each date. This is the baseline S-curve. It ends at the budget at completion (BAC).
• Earned value (EV): the cumulative budget for the work actually done. It is plotted only up to the status date.
• Actual cost (AC): the cumulative money actually spent on that work, also up to the status date.

The gaps between the lines at the status date carry the message:

• EV below PV: less work done than planned, so behind schedule. Schedule variance SV = EV - PV is negative.
• AC above EV: spent more than the work is worth, so over budget. Cost variance CV = EV - AC is negative.
• AC below PV on its own means nothing. Lower spending can mean efficiency or simply that work has not been done. This is the classic misreading of a two-line chart, and the reason EV is needed.

The earned value management guide covers the full set of measures.

How to build a project S-curve

1. Start from a costed, logic-linked schedule. Each activity needs dates and a budget.
2. Spread each activity's budget over its duration. Evenly is the simple default; front-loaded or back-loaded profiles fit some work better (materials bought at the start, payment on completion).
3. Sum the spread by period, usually by week or month, across all activities.
4. Accumulate. Each period's cumulative value is the previous total plus this period's amount. That running total is your PV S-curve.
5. Freeze it as the baseline. Then add EV and AC each reporting period.

Some teams also draw an early-start curve and a late-start curve. The early curve assumes every activity starts as soon as logic allows; the late curve assumes every activity uses all its float. The band between them shows how much room the plan has. Actual progress below the late curve means the project is eating into the critical path.

Worked example: reading an S-curve at month 3

A six-month project has a BAC of $600,000. The baseline monthly spend and its running total are:

Month	1	2	3	4	5	6
Planned spend ($000)	40	80	140	160	120	60
Cumulative PV ($000)	40	120	260	420	540	600

The monthly figures rise and fall like a hill; the cumulative row traces the S. At the end of month 3, the status is EV = $220,000 and AC = $250,000, against PV = $260,000.

• SV = 220,000 - 260,000 = -$40,000, and SPI = 220,000 / 260,000 = 0.85 (to two decimals). Behind schedule.
• CV = 220,000 - 250,000 = -$30,000, and CPI = 220,000 / 250,000 = 0.88. Over budget.

Now read the curve horizontally instead of vertically. The EV of $220,000 is where the PV curve stood partway through month 3. PV was 120 at the end of month 2 and 260 at the end of month 3, so 220 is reached at 2 + (220 - 120) / (260 - 120) = 2 + 100 / 140 = 2.71 months. That is the earned schedule: the project has done the work planned for 2.71 months in 3 months. SPI(t) = 2.71 / 3 = 0.90, and a simple duration forecast is 6 / 0.905 = about 6.6 months. The horizontal gap on an S-curve is the schedule delay in time units, which is easier to explain than a dollar figure. The earned schedule calculator does this interpolation for you.

Common S-curve mistakes

• Plotting only planned and actual cost. Without EV, you cannot tell underspending from slow progress.
• Re-drawing the baseline every month. If the PV curve moves to meet actuals, it stops measuring anything. Change it only through approved changes, and keep the old baseline for comparison.
• Spreading cost evenly when it is not even. A large purchase spread across six months creates false variances every month.
• Trusting SV after the planned finish. When the project runs late, PV stays flat at BAC after the planned finish, which makes SV shrink toward zero even though the project is still late. Earned schedule avoids this problem.
• Ignoring the slope. If the actual curve must become much steeper than anything achieved so far to reach the finish, the forecast date is not credible.

How to do this in Critova

Critova holds the inputs an S-curve needs: a critical path schedule with baselines, cost entries entered by hand or imported from a spreadsheet, and progress recorded by the team. From these it calculates PV, EV and AC, the variances, CPI, SPI and the earned schedule measure SPI(t), each with its formula, and the cost and schedule dashboards bring them together. If you want to plot the curve yourself in a spreadsheet, export the cost and activity registers to Excel. See the cost and earned value features.