Lead mid-size IT projects: hybrid delivery, governance, earned value, vendors, business cases, technical and security risk, recovery and your path to PMP and program management.

IT Project Leadership

## Why it matters

"We have spent 42% of the budget" sounds like useful information, but on its own it tells you almost nothing. Is 42% spent good or bad? It depends on how much work is done and how much was supposed to be done by now. **Earned value management (EVM)** combines scope, schedule and cost into a few numbers that answer two questions every sponsor asks: *are we on schedule?* and *are we on budget?* It is used widely on government and large IT projects, and its indices appear in PMI's certification material.

## The three base numbers

Pick a status date, for example the end of month 4.

- **Planned value (PV)**: the budgeted cost of the work that was *scheduled* to be done by the status date. It comes from your baseline plan.
- **Earned value (EV)**: the budgeted cost of the work *actually completed* by the status date. If the whole project is budgeted at $480,000 and 37.5% of the work is done, EV = $180,000.
- **Actual cost (AC)**: what the completed work *actually cost*, from timesheets and invoices.

The whole project's budget is the **budget at completion (BAC)**.

Earned value is the clever part: it measures progress in money, using the *budget* for the work done, so it can be compared directly with both the plan and the spend.

## Variances and indices

| Measure | Formula | Meaning |
|---|---|---|
| Cost variance | CV = EV - AC | Negative = over budget for the work done |
| Schedule variance | SV = EV - PV | Negative = behind schedule (in money terms) |
| Cost performance index | CPI = EV / AC | Value earned per dollar spent; below 1 = over budget |
| Schedule performance index | SPI = EV / PV | Rate of progress vs plan; below 1 = behind |

Indices are often more useful than variances because they are ratios: a CPI of 0.9 means the same thing on a $50,000 project and a $5 million one.

## Forecasting

If the cost efficiency so far is likely to continue, a common forecast is:

```text
EAC = BAC / CPI          estimate at completion
ETC = EAC - AC           estimate to complete (what is still to spend)
VAC = BAC - EAC          variance at completion
```

There are other EAC formulas (for example, if you believe the overrun was a one-off, EAC = AC + (BAC - EV)). Say which one you used.

## Worked example: Northside Clinics at month 4

The baseline says that by the end of month 4 the project should have done $240,000 worth of work. The team estimates 37.5% of the work is complete, and finance reports $200,000 spent.

```text
BAC = 480,000
PV  = 240,000
EV  = 0.375 x 480,000 = 180,000
AC  = 200,000

CV  = 180,000 - 200,000 = -20,000   (over budget for the work done)
SV  = 180,000 - 240,000 = -60,000   (behind plan)
CPI = 180,000 / 200,000 = 0.90
SPI = 180,000 / 240,000 = 0.75
EAC = 480,000 / 0.90    = 533,333
VAC = 480,000 - 533,333 = -53,333
```

In plain English for the steering committee:

> "We are getting 90 cents of planned work for every dollar spent, and we have done three quarters of the work we planned by now. If nothing changes, the project will cost about $533,000, $53,000 over budget. The main cause is the integration delay; the recovery options are on page 2."

Notice that spend alone ($200,000, below the planned $240,000) would have looked like an *underspend*. EVM shows the project is in fact behind and over budget for what it has delivered.

## Measuring "percent complete" honestly

EV is only as good as your progress measure. Guessing "about 40%" invites optimism. Better rules:

- **0/100**: a work package earns nothing until it is done. Strict and simple for small packages.
- **50/50**: half when started, half when finished.
- **Milestone weights**: for example design 20%, build 50%, tested 30%.
- **Agile projects**: use completed story points against the release plan (accepted stories only, not "nearly done").

## Common mistakes

- Treating "money spent" as progress.
- Mixing up the formulas: in both indices, **EV is on top**. CPI uses AC; SPI uses PV.
- Reporting CPI to three decimals as if it were precise when the percent complete was a guess.
- Ignoring that SPI drifts back to 1.0 near the end of a late project (EV eventually equals PV once everything is done), so check the critical path too.

## Quick recap

- PV = planned work, EV = work done (at budget rates), AC = actual cost.
- CPI = EV / AC and SPI = EV / PV; below 1 is bad.
- EAC = BAC / CPI forecasts the final cost if efficiency stays the same.
- Measure progress with clear rules, not optimism.

## Try it

Open the **Do** tab. The earned value calculator is written except for one line: the CPI formula. Fill it in and the program produces the month 4 report. No programming experience needed.


Use earned value management to compare planned value, earned value and actual cost, and compute CPI, SPI and a cost forecast.

Budgets and earned value

What is Earned Value Management? Example Formulas | EV, PV, SV, CV, AC

Cost Performance Index and Schedule Performance Index - CPI and SPI Example

## Why it matters

Few IT projects are delivered entirely in-house. You buy software licences, cloud services, integration work, hardware and specialist consultants. Every one of those comes with a contract, and the contract type decides who carries the risk when things change. A project lead does not need to be a lawyer, but you must understand the main contract types, what a good statement of work contains, and how to manage a vendor day to day. Always involve your procurement and legal teams: they own the contract, and you own delivery.

## The main contract types

**Fixed price (FP)**. The vendor agrees to deliver a defined scope for a set price. If it takes them longer, they absorb the cost.

- Good for the buyer when scope is clear and stable.
- The vendor adds a risk premium to the price, so you pay for certainty.
- Every change in scope becomes a **change order** with a new price. On uncertain work this can become slow and expensive.
- Variants include fixed price with incentive fees and fixed price with economic price adjustment (for long contracts affected by inflation).

**Time and materials (T&M)**. You pay an agreed rate per hour or day, plus materials, for the time actually worked.

- Good when scope is uncertain or likely to change, such as an agile build or a discovery phase.
- The buyer carries the cost risk: if work takes longer, you pay more.
- Protect yourself with a **not-to-exceed (NTE) cap**, regular burn reports and the right to review timesheets.

**Cost-reimbursable (cost-plus)**. The vendor is paid their actual costs plus a fee (fixed, or linked to performance). Common in research and large government programmes where the work cannot be defined up front. The buyer carries most of the risk and needs strong cost oversight.

## Who carries the risk?

| Contract type | Buyer cost risk | Flexibility to change scope | Best when |
|---|---|---|---|
| Fixed price | Low | Low (change orders) | Scope is well defined |
| Time and materials | High (cap it) | High | Scope is uncertain or evolving |
| Cost-reimbursable | Highest | High | Work cannot be defined; research |

A rule of thumb: **the party with the most control over the uncertainty should carry the risk**. If the buyer keeps changing its mind, a fixed-price contract will be expensive and argumentative.

## Worked example: two vendors on Northside Clinics

1. **Records integration.** The records vendor publishes a stable API, and the work is to build and test six defined interfaces. Northside uses a **fixed-price** contract with payment tied to milestones: 30% on design approval, 50% on successful integration testing, 20% after 30 days of stable production use.
2. **Portal user research and design.** Nobody knows yet what patients will need, and the team expects to iterate. Northside uses **T&M** at an agreed day rate with a not-to-exceed cap of $60,000 and a monthly burn report. If the cap is approached, the sponsor decides whether to extend.

A hybrid project often ends up with a hybrid set of contracts, which mirrors the delivery choices from module 1.

## The statement of work

The **statement of work (SOW)** describes what the vendor will do. It often sits under a **master services agreement (MSA)** that holds the legal terms. A good SOW includes:

- Objectives and background.
- Scope: deliverables, and explicit **exclusions**.
- Acceptance criteria: how you will decide a deliverable is done.
- Milestones and payment schedule.
- Roles, responsibilities and what the buyer must provide (access, test data, decisions within five days).
- Service levels (an **SLA**) for ongoing services: availability, response times, and what happens if they are missed.
- Change control: how changes are requested, priced and approved.

Most vendor disputes trace back to vague acceptance criteria or buyer responsibilities that were never written down.

## Managing vendors day to day

- Treat the vendor team as part of the project: include them in planning, risk reviews and demos.
- Track deliverables against the SOW, not against what was "agreed on a call". Confirm verbal agreements in writing.
- Meet your own obligations on time. A late test environment from your side gives the vendor a fair claim for delay.
- Escalate through the agreed route (account manager, then contract owner), with facts.
- Review invoices against milestones or timesheets before approving them.

## Common mistakes

- Choosing fixed price for work nobody can define yet, then fighting over every change order.
- T&M with no cap and no burn reporting.
- Acceptance criteria such as "to the client's satisfaction".
- Letting the vendor's project manager run your project.

## Quick recap

- Fixed price moves cost risk to the vendor but needs clear scope; T&M is flexible but the buyer carries the cost risk; cost-plus is for undefinable work.
- The party best able to control the uncertainty should carry the risk.
- A SOW defines scope, exclusions, acceptance, milestones, responsibilities and change control.
- Manage vendors as part of the team, in writing, and meet your own obligations.

## Try it

Open the **Do** tab and test yourself on contract types and statements of work.


Compare fixed-price, time-and-materials and cost-reimbursable contracts, and manage vendors through a clear statement of work.

Vendors and contracts

Fixed-Price Contracts VS Time-and-Materials for Your Software Development Project

Vendor Management: How to Work with Vendors

Write the problem in two or three sentences with at least one measured baseline (time, cost, errors, complaints).

Describe three options, one of which is do nothing or do minimum.

For each option, list one-off costs and yearly running costs. Ask someone or look up public prices to make them realistic.

List the benefits of your preferred option. For each one, give a yearly value (or say it is non-financial), a benefit owner and how it will be measured.

Build a three-year table of costs, benefits, net and cumulative net. Calculate ROI = (total benefits - total costs) / total costs and the payback year.

Do one sensitivity check: halve your biggest benefit and recalculate ROI and payback.

Write a one-paragraph recommendation that states the result honestly, including the sensitivity.

## Why it matters

Every project competes for money and people with other projects. The **business case** is the document that justifies the investment: why do this, why now, why this option, and what will it cost and return? In many organisations the project lead helps the sponsor write it, and then uses it all the way through the project to check the investment still makes sense (that is what stage gates check). An IT project manager who can talk about value, not just tasks, earns a seat in better conversations.

## What a business case contains

1. **Problem or opportunity.** What is wrong today, with evidence.
2. **Strategic fit.** Which organisational goal it supports.
3. **Options.** At least three, including **do nothing** (or "do minimum"). The do-nothing option is your baseline: it has costs too.
4. **Costs** for each option: one-off (build, licences, hardware, training) and ongoing (subscriptions, support, staff).
5. **Benefits**: measurable ones (money saved or earned, hours freed), and non-financial ones (patient experience, compliance, risk reduction). Name a benefit owner for each.
6. **Financial summary**: ROI, payback period and, for larger investments, net present value.
7. **Risks and assumptions.**
8. **Recommendation.**

## The key numbers

**Return on investment (ROI)** over a stated period:

```text
ROI = (total benefits - total costs) / total costs
```

**Payback period**: how long until cumulative benefits equal cumulative costs.

**Net present value (NPV)** recognises that money in the future is worth less than money today. Each year's net cash flow is discounted by the organisation's discount rate *r*:

```text
NPV = sum over years t of  net cash flow in year t / (1 + r)^t
```

A positive NPV means the investment returns more than the discount rate. Your finance team will usually tell you which rate and which measures they expect; ask them early.

## Worked example: Northside Clinics

**Problem.** Phone lines are swamped at 8 a.m.: 22% of callers hang up before reaching anyone (from the phone system's reports), and 9% of appointments are missed because there are no reminders.

**Options.**

- **A. Do nothing.** Keep phones only. Costs continue: overtime for call handlers, and lost revenue from missed appointments.
- **B. Add call handlers.** Hire four more staff: $200,000 a year. Improves answer rates but not missed appointments.
- **C. Online booking portal with reminders** (the recommended option): $480,000 to build over nine months, then $60,000 a year to run.

**Benefits of option C** (each with an owner, the operations director):

- Reduce call-handling overtime: $90,000 a year.
- Cut missed appointments from 9% to 5% using text reminders: about $150,000 a year in recovered appointment revenue.
- Non-financial: better patient access, fewer complaints, retiring 300 shared logins (security).

**Three-year view** (benefits start in year 1 at half value, because rollout finishes mid-year):

| | Year 0 | Year 1 | Year 2 | Year 3 |
|---|---|---|---|---|
| Costs | 480,000 | 60,000 | 60,000 | 60,000 |
| Benefits | 0 | 120,000 | 240,000 | 240,000 |
| Net | -480,000 | 60,000 | 180,000 | 180,000 |
| Cumulative | -480,000 | -420,000 | -240,000 | -60,000 |

Over three years: total benefits $600,000, total costs $660,000, so ROI = (600,000 - 660,000) / 660,000 = **-9%**. Payback comes in year 4 (cumulative turns positive when year 4 adds another $180,000).

That is an honest result, and it is useful. The sponsor might decide that a five-year view is fairer for a system expected to last eight years (five-year ROI is about +38%), or that the non-financial benefits justify it anyway. A good business case shows the numbers plainly and lets the decision-makers decide. It never stretches benefits to reach a target.

## Making benefits credible

- Use a **baseline** measured today ("22% abandoned calls in Q1"), not a guess.
- Be conservative and show a range or a sensitivity: "if missed appointments only fall to 7%, the missed-appointment benefit drops to $75,000 a year and payback moves to year 6."
- Separate **cash-releasing** benefits (money actually saved) from **efficiency** benefits (hours freed that may be used elsewhere).
- Plan how each benefit will be measured after go-live, and by whom.

## Common mistakes

- No do-nothing option, so there is no baseline.
- Counting only build costs and forgetting running costs, support and training.
- Optimistic benefits with no owner.
- Writing the business case once to get approval, then never looking at it again.

## Quick recap

- A business case compares options, including do nothing, on costs, benefits and risks.
- ROI = (benefits - costs) / costs over a stated period; payback is when cumulative net turns positive; NPV discounts future cash.
- Credible benefits have baselines, owners, ranges and a measurement plan.
- Revisit the business case at every stage gate.

## Try it

Open the **Do** tab and write a short business case for a project you know.


Write a business case that compares options and shows costs, benefits, ROI and payback, so sponsors can decide whether to invest.

Business cases

What is a Business Case? Project Management in Under 5

How to Write a Business Case - Project Managment Training

Track cost and schedule with earned value, choose the right vendor contract, and write a business case with ROI and payback.