$560 million. That's what Denver International Airport spent in the nineties on a fully automated baggage handling system – two years late, double the budget, and it never really worked. The problem wasn't technology. It was over 2,000 change requests from stakeholders that nobody managed properly. Sounds absurd – until you open your own agency's Slack and see the client dropping in "just one small thing" at 4:47 PM on a Friday.

IT project management is one of those topics where most people assume they're doing fine. The numbers tell a different story.

IT project management – beyond Gantt charts and status meetings

The Standish Group has published the CHAOS Report since 1994, one of the largest longitudinal studies on IT projects worldwide. The 2020 results: 31% of all IT projects succeed. 50% are classified as "challenged" – they cross the finish line somehow, but late, over budget, or with reduced scope. 19% fail outright.

31% success. 50% limping across the line. 19% total failure. And that's after thirty years of improvement – in 1994, the success rate was 16%.

For IT service providers and agencies, the challenge is even sharper. Unlike internal IT departments, service firms run 5, 10, sometimes 20 projects in parallel. Each with a different client, different expectations, different communication channels. IT project management in this context doesn't mean shepherding one initiative through a phase model. It means steering a portfolio of live projects so there's still margin left at the end of the month.

This is where the textbook falls apart. In the textbook, a project has one sponsor, one team, one budget. In an agency, three projects share the same developer, the sponsor is simultaneously the client's intern, and the budget was "slightly optimistic" during the sales pitch.

What it's actually about

IT project management for service providers boils down to keeping three things in balance at the same time: client expectations, team capacity, and the financial health of the project. Manage only two and you'll eventually have a problem with the third. Happy clients at negative margins? Works for a quarter, then you're bankrupt. Great margins with an overloaded team? Lasts until your best people quit. The triangle has to hold.

Agile, Scrum, Waterfall – which method fits whom?

The methodology debate has filled conference stages and LinkedIn feeds for twenty years. The short version: agile projects have a 42% success rate according to the 2020 CHAOS Report. Waterfall sits at 13%. Seems clear-cut – but it isn't.

The honest answer for most IT service providers: you're already working hybrid – you just don't call it that. Almost nobody runs pure Scrum with all the ceremonies. The Scrum Master is usually called "project manager" and secretly maintains a Gantt chart because the client wants to see one. And almost nobody runs pure Waterfall, because requirements change before the specification is even finished.

The method is rarely the problem

I've seen agencies that run pure Waterfall profitably – because they're exceptionally good at requirements gathering. And I've seen agencies that claim to be "agile" and still tank every other project – because "agile" means there's no planning.

The real question isn't which methodology you follow. It's whether your team has a shared understanding of how you run projects at all. If every project manager is running their own system – in Excel, Notion, Jira, or in their head – the methodology is irrelevant. You don't have IT project management. You have organized chaos with individual workarounds.

If you're curious about the tension between agile methods and rigorous time tracking, our piece on agile time tracking covers some uncomfortable truths.

The 7 most expensive mistakes in IT project management

I could claim we never made any of these mistakes at Leadtime. That would be a lie. Here are the seven I've seen most often over 15 years – in our own company and across hundreds of agencies we've talked to.

1. Scope creep without a control mechanism. PMI measured it in 2018: 52% of all projects experience uncontrolled scope changes. Organizations with structured change management cut that to 33%. The difference isn't complicated processes – it's someone asking: "Is this in scope? If not, what does the change cost and who approves it?" One sentence that saves thousands.

2. Treating time tracking as a billing chore instead of a steering tool. Most agencies track time so the invoice is correct. Few use the data in real time to spot that Project X has been eating margin for three weeks. Discovering on Friday at 5 PM that the budget is gone isn't controlling – it's archaeology. If you treat project time tracking as a pure billing function, you're missing the point entirely.

3. Resource planning by gut feeling. "Tobias probably has capacity next week." He doesn't. But you won't find out until Monday, when three projects ask for him simultaneously. The Standish Group found a clear correlation: small projects with dedicated teams succeed around 90% of the time. Large projects with shared resources drop below 10%. Team size alone doesn't explain the gap – clarity of assignment does.

4. No unified project tool. Every project manager uses their favorite tool. One swears by Trello, another by Asana, a third does everything in Excel. Result: nobody has a holistic view across all running projects. If you're steering more than five projects at once, you can't avoid the topic of multi-project management – and the uncomfortable realization that tool diversity isn't a feature, it's a risk.

5. Fixed price without a buffer. Fixed-price projects work – if you build in 20–30% buffer and have an honest conversation with the client about it. Most agencies budget 5% and then wonder why every third project goes red.

6. Status meetings without substance. Weekly "what did you do last week?" meetings where everyone reads their to-do list. Costs 10–20 person-hours per month depending on team size. A good dashboard replaces 80% of these meetings. The remaining 20% become actual decision sessions – instead of mandatory events where everyone reads emails on the side.

7. Projects never officially end. No retro, no lessons learned, no clean close. The project just evaporates – the last task gets done, the invoice goes out, done. And six months later you make the same mistakes in the next project.

IT project management software: a practical comparison

The tool landscape for IT project management software is overwhelming – and changes faster than any comparison article can keep up with. Instead of a feature shoot-out that'll be outdated next month, here are the four categories and their respective blind spots.

The real question isn't: which tool has the most features? It's: which tool tells you on Friday evening whether your projects are still profitable?

The Frankenstein stack

Most IT service providers end up with a grown stack: Jira for development, Harvest or Toggl for time tracking, Excel for budgets, Slack for communication. Each tool works fine on its own. Together they create data silos – and nobody has the full picture.

The issue isn't that individual tools are bad. It's the gap between them. The developer tracks time in Tool A. The project manager plans in Tool B. Leadership checks finances in Tool C. By the time the numbers are in sync, the week is over. IT project management tools that don't close these silos just move the problem around.

What service providers actually need

A tool for IT project management in agencies needs to do four things that classic PM software doesn't cover:

First, link time tracking and project budget in real time – not at the end of the month when invoices go out. Second, make utilization visible across all projects, not per project but per person and per week. Third, attach client communication and support tickets to the project, not hide them in a separate system. Fourth, treat finances – hourly rates, margins, contribution margins – as a natural part of project management, not as an accounting afterthought.

Why classic PM tools aren't enough for IT service providers

Here's where it gets fundamental. Most PM tools were built for internal project teams: one project, one team, one budget. IT service providers live in a different reality – many projects running in parallel, shared resources, external clients with expectations that shift weekly. That reality doesn't fit into a Gantt chart.

Classic IT project management software thinks in tasks and milestones. Service providers think in hours and margins. If your PM tool can't tell you whether Project X will still be profitable at the current burn rate – it's a task manager with a nice interface. Not a steering instrument.

The gap sits between two worlds: the Jira world (tickets, sprints, velocity) and the ERP world (invoicing, controlling, contribution margins). Both have their place. But most service providers stand somewhere in between, bridging the gap with spreadsheets that nobody keeps current.

We built Leadtime for exactly this gap – project management, ticketing, and finances in one system. Not because we were smarter than everyone else, but because we ran exactly this Frankenstein stack at our previous company for years and eventually got tired of copying numbers between five tools.

IT project management doesn't get better through the right methodology. Or the right tool. It gets better when you stop treating method, tooling, and reality as three separate worlds – and start thinking of them as one system.