Schedule Variance Percentage (SV%) normalises Schedule Variance so you can compare programme performance across projects of different sizes. Raw SV is measured in pounds, which makes cross-project comparison meaningless. A -£500K SV on a £5M project is a crisis. The same -£500K on a £50M project is a rounding error. SV% strips out the scale and tells you how far behind (or ahead) you are as a percentage of the plan.
SV% = (SV / PV) x 100
Or equivalently: SV% = ((EV - PV) / PV) x 100
This term is part of the earned value definitions glossary. For the raw pound-value metric, see Schedule Variance (SV).
The Formula
SV% = SV / PV x 100Where:
- SV = Earned Value minus Planned Value
- PV = cumulative budgeted cost of work scheduled to date
| SV% Range | Status | Typical Response |
|---|---|---|
| > +10% | Well ahead | Verify it's genuine progress, not front-loading |
| +5% to +10% | Ahead | Monitor. Ensure quality isn't being sacrificed for speed |
| -5% to +5% | On track | Normal variance band on most construction projects |
| -10% to -5% | Slipping | Investigate root cause. Yellow flag |
| < -10% | Serious delay | Mandatory recovery plan. Red flag |
These thresholds aren't gospel. I've worked on projects where -5% was acceptable (complex M&E with long procurement lead times) and others where -3% triggered an escalation (time-critical rail possessions with fixed handback dates). Set your thresholds during baseline approval and stick to them.
Why SV% Matters: The Normalisation Problem
Here's the scenario that makes SV% essential.
WHY RAW SV IS MISLEADING ACROSS PROJECTS ============================================= PROJECT ALPHA (£5M retail fit-out) ├── PV at month 4 = £2,000,000 ├── EV at month 4 = £1,500,000 ├── SV = -£500,000 └── SV% = -£500K / £2M = -25% ← CRISIS. Quarter of planned work not done. PROJECT BETA (£50M hospital) ├── PV at month 10 = £22,000,000 ├── EV at month 10 = £21,500,000 ├── SV = -£500,000 └── SV% = -£500K / £22M = -2.3% ← Minor blip. Easily recoverable. ───────────────────────────────────────────── SAME SV (-£500K) COMPLETELY DIFFERENT SEVERITY ───────────────────────────────────────────── Without SV%, a portfolio director looking at two SV figures of -£500K might treat both projects equally. With SV%, it's obvious: Alpha is in serious trouble, Beta is fine.
On a programme with 8 or 10 live projects, the portfolio director doesn't have time to mentally adjust every SV figure for project size. SV% does that automatically. It's the metric that belongs in the programme-level dashboard.
The Relationship Between SV%, SV, and SPI
These three metrics all measure the same thing, schedule performance, but from different angles:
THREE WAYS TO SAY "WE'RE BEHIND SCHEDULE" ============================================= SV = EV - PV = -£1,300,000 (how much behind, in pounds) SV% = SV / PV x 100 = -15.5% (how much behind, as a proportion of plan) SPI = EV / PV = 0.845 (efficiency: £0.85 of progress per £1 planned) ───────────────────────────────────────────── MATHEMATICAL RELATIONSHIP: SV% = (SPI - 1) x 100 ───────────────────────────────────────────── So SPI = 0.845 → SV% = (0.845 - 1) x 100 = -15.5% Use SV for absolute impact ("we're £1.3M behind") Use SV% for cross-project comparison ("Project A is -15%, B is -3%") Use SPI for efficiency and forecasting ("at this rate.")The mathematical link between SV% and SPI is straightforward: SV% = (SPI - 1) x 100. An SPI of 0.90 always gives SV% = -10%. They're two expressions of the same ratio.
Worked Example: Comparing Two Projects in a Programme
Worked ExampleScenario: A Tier 1 contractor runs two NEC4 Option C projects simultaneously. The commercial director reviews both at the monthly programme board (October 2025).
Project 1: £5M Retail Fit-out in Manchester
- BAC = £5,000,000
- Duration: 8 months
- Current month: 4 of 8 (50% through programme)
- PV = £2,000,000
- EV = £1,500,000
- SV = £1,500,000 - £2,000,000 = -£500,000
- SV% = -£500,000 / £2,000,000 x 100 = -25.0%
- SPI = 0.750
Project 2: £50M Hospital Extension in Birmingham
- BAC = £50,000,000
- Duration: 24 months
- Current month: 10 of 24 (42% through programme)
- PV = £22,000,000
- EV = £21,500,000
- SV = £21,500,000 - £22,000,000 = -£500,000
- SV% = -£500,000 / £22,000,000 x 100 = -2.3%
- SPI = 0.977
The dashboard without SV%:
| Project | SV |
|---|---|
| Retail fit-out | -£500,000 |
| Hospital extension | -£500,000 |
Looks equal. Feels like both need the same intervention.
The dashboard with SV%:
| Project | SV | SV% | SPI | Status |
|---|---|---|---|---|
| Retail fit-out | -£500,000 | -25.0% | 0.750 | RED. Recovery plan needed immediately |
| Hospital extension | -£500,000 | -2.3% | 0.977 | GREEN. Normal variance, monitor |
Now the commercial director knows exactly where to focus. The retail fit-out at -25% needs a recovery plan by Friday. The hospital extension at -2.3% just needs monitoring.
Common Mistakes
- Using SV instead of SV% for portfolio reporting. If you're comparing multiple projects, raw SV is misleading. A £60M project will almost always have a larger SV than a £3M project. SV% lets you compare apples to apples.
- Setting the same SV% threshold for all project types. A -10% threshold makes sense for a straightforward civils job. On a complex M&E-heavy project with long procurement chains, -10% might be structurally normal for the first three months while equipment is on order. Tailor thresholds to the project type.
- Forgetting the convergence problem. SV% inherits the same convergence-to-zero flaw as SV. As the project nears completion, both EV and PV approach BAC, so SV shrinks and SV% trends towards zero. In the last 20% of a project, SV% becomes unreliable. Use time-based metrics like SV(t) instead.
- Dividing by zero at the start. At month 0, PV = 0. SV% = SV / 0 is undefined. Most EVM systems handle this by returning null until the first reporting period where PV > 0. Don't manually force it to zero. That implies on-track performance when you have no data.
How Gather helps. Gather's AI reads your site diaries daily and maps progress against your cost-loaded programme, giving you accurate earned value data without manual spreadsheet updates. Book a demo to see it working on a live NEC4 project.
Frequently Asked Questions
What SV% is considered acceptable on UK construction projects?
Most Tier 1 contractors I've worked with use a traffic light system: green at -5% to +5%, amber at -10% to -5%, red below -10%. But these aren't industry standards, they're project-specific thresholds set during baseline approval. On a fast-track programme with tight possession windows, even -5% might trigger escalation. On a long-duration infrastructure scheme, -8% in the early months might be acceptable if the critical path has float.
How often should SV% be calculated?
Monthly, aligned with your earned value reporting cycle. Some projects calculate it fortnightly, particularly on shorter-duration contracts (under 6 months) where a monthly snapshot doesn't give enough granularity. Weekly is usually overkill unless you're in a recovery phase and need to track whether acceleration measures are working.
Does SV% suffer from the same convergence problem as SV?
Yes. SV% = SV / PV, and since SV converges to zero at completion, SV% does too. A project that finishes 4 months late will show SV% = 0 at completion. For the second half of any project, supplement SV% with SPI(t) or SV(t) for an honest picture of schedule performance.
.webp)
